WO2021249329A1 - 一种分子级装置的实时优化方法、装置、系统及存储介质 - Google Patents

一种分子级装置的实时优化方法、装置、系统及存储介质 Download PDF

Info

Publication number
WO2021249329A1
WO2021249329A1 PCT/CN2021/098570 CN2021098570W WO2021249329A1 WO 2021249329 A1 WO2021249329 A1 WO 2021249329A1 CN 2021098570 W CN2021098570 W CN 2021098570W WO 2021249329 A1 WO2021249329 A1 WO 2021249329A1
Authority
WO
WIPO (PCT)
Prior art keywords
product
group
single molecule
predicted
preset
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2021/098570
Other languages
English (en)
French (fr)
Inventor
王杭州
纪晔
刘一心
孙宝文
石振民
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Petrochina Co Ltd
Original Assignee
Petrochina Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Petrochina Co Ltd filed Critical Petrochina Co Ltd
Priority to EP21822840.1A priority Critical patent/EP4089680B1/en
Priority to BR112022017194A priority patent/BR112022017194A2/pt
Publication of WO2021249329A1 publication Critical patent/WO2021249329A1/zh
Priority to US18/047,036 priority patent/US20230110441A1/en
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16CCOMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C10/00Computational theoretical chemistry, i.e. ICT specially adapted for theoretical aspects of quantum chemistry, molecular mechanics, molecular dynamics or the like
    • CCHEMISTRY; METALLURGY
    • C10PETROLEUM, GAS OR COKE INDUSTRIES; TECHNICAL GASES CONTAINING CARBON MONOXIDE; FUELS; LUBRICANTS; PEAT
    • C10GCRACKING HYDROCARBON OILS; PRODUCTION OF LIQUID HYDROCARBON MIXTURES, e.g. BY DESTRUCTIVE HYDROGENATION, OLIGOMERISATION, POLYMERISATION; RECOVERY OF HYDROCARBON OILS FROM OIL-SHALE, OIL-SAND, OR GASES; REFINING MIXTURES MAINLY CONSISTING OF HYDROCARBONS; REFORMING OF NAPHTHA; MINERAL WAXES
    • C10G11/00Catalytic cracking, in the absence of hydrogen, of hydrocarbon oils
    • C10G11/14Catalytic cracking, in the absence of hydrogen, of hydrocarbon oils with preheated moving solid catalysts
    • C10G11/18Catalytic cracking, in the absence of hydrogen, of hydrocarbon oils with preheated moving solid catalysts according to the "fluidised-bed" technique
    • C10G11/187Controlling or regulating
    • CCHEMISTRY; METALLURGY
    • C10PETROLEUM, GAS OR COKE INDUSTRIES; TECHNICAL GASES CONTAINING CARBON MONOXIDE; FUELS; LUBRICANTS; PEAT
    • C10GCRACKING HYDROCARBON OILS; PRODUCTION OF LIQUID HYDROCARBON MIXTURES, e.g. BY DESTRUCTIVE HYDROGENATION, OLIGOMERISATION, POLYMERISATION; RECOVERY OF HYDROCARBON OILS FROM OIL-SHALE, OIL-SAND, OR GASES; REFINING MIXTURES MAINLY CONSISTING OF HYDROCARBONS; REFORMING OF NAPHTHA; MINERAL WAXES
    • C10G35/00Reforming naphtha
    • C10G35/24Controlling or regulating of reforming operations
    • CCHEMISTRY; METALLURGY
    • C10PETROLEUM, GAS OR COKE INDUSTRIES; TECHNICAL GASES CONTAINING CARBON MONOXIDE; FUELS; LUBRICANTS; PEAT
    • C10GCRACKING HYDROCARBON OILS; PRODUCTION OF LIQUID HYDROCARBON MIXTURES, e.g. BY DESTRUCTIVE HYDROGENATION, OLIGOMERISATION, POLYMERISATION; RECOVERY OF HYDROCARBON OILS FROM OIL-SHALE, OIL-SAND, OR GASES; REFINING MIXTURES MAINLY CONSISTING OF HYDROCARBONS; REFORMING OF NAPHTHA; MINERAL WAXES
    • C10G45/00Refining of hydrocarbon oils using hydrogen or hydrogen-generating compounds
    • C10G45/72Controlling or regulating
    • CCHEMISTRY; METALLURGY
    • C10PETROLEUM, GAS OR COKE INDUSTRIES; TECHNICAL GASES CONTAINING CARBON MONOXIDE; FUELS; LUBRICANTS; PEAT
    • C10GCRACKING HYDROCARBON OILS; PRODUCTION OF LIQUID HYDROCARBON MIXTURES, e.g. BY DESTRUCTIVE HYDROGENATION, OLIGOMERISATION, POLYMERISATION; RECOVERY OF HYDROCARBON OILS FROM OIL-SHALE, OIL-SAND, OR GASES; REFINING MIXTURES MAINLY CONSISTING OF HYDROCARBONS; REFORMING OF NAPHTHA; MINERAL WAXES
    • C10G47/00Cracking of hydrocarbon oils, in the presence of hydrogen or hydrogen- generating compounds, to obtain lower boiling fractions
    • C10G47/36Controlling or regulating
    • CCHEMISTRY; METALLURGY
    • C10PETROLEUM, GAS OR COKE INDUSTRIES; TECHNICAL GASES CONTAINING CARBON MONOXIDE; FUELS; LUBRICANTS; PEAT
    • C10GCRACKING HYDROCARBON OILS; PRODUCTION OF LIQUID HYDROCARBON MIXTURES, e.g. BY DESTRUCTIVE HYDROGENATION, OLIGOMERISATION, POLYMERISATION; RECOVERY OF HYDROCARBON OILS FROM OIL-SHALE, OIL-SAND, OR GASES; REFINING MIXTURES MAINLY CONSISTING OF HYDROCARBONS; REFORMING OF NAPHTHA; MINERAL WAXES
    • C10G9/00Thermal non-catalytic cracking, in the absence of hydrogen, of hydrocarbon oils
    • C10G9/005Coking (in order to produce liquid products mainly)
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16CCOMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/20Identification of molecular entities, parts thereof or of chemical compositions
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16CCOMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/90Programming languages; Computing architectures; Database systems; Data warehousing
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16CCOMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C60/00Computational materials science, i.e. ICT specially adapted for investigating the physical or chemical properties of materials or phenomena associated with their design, synthesis, processing, characterisation or utilisation
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16CCOMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/10Analysis or design of chemical reactions, syntheses or processes
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16CCOMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/30Prediction of properties of chemical compounds, compositions or mixtures
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16CCOMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/70Machine learning, data mining or chemometrics

Definitions

  • the present invention relates to the technical field of petroleum processing, in particular to a real-time optimization method, device, system and storage medium of a molecular-level device.
  • the present invention provides a method, device, system and storage medium for real-time optimization of molecular-level devices.
  • the present invention provides a method for real-time optimization of a molecular-level device, which includes the following steps:
  • the corresponding fractions are used as the petroleum processing raw materials, respectively, and input the pre-trained product prediction model corresponding to the petroleum processing device to obtain the predicted molecular composition and the predicted molecular composition of the corresponding predicted product output by the product prediction model.
  • the predicted molecular composition of the predicted product and the predicted molecular content of each single molecule in the predicted molecular composition it is determined whether the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set ;
  • the predicted product does not meet any of the preset standards of the target product corresponding to the predicted product in the preset standard set, then adjust the operating parameters in the product prediction model to reacquire the predicted product
  • the predicted molecular composition and the predicted molecular content of each single molecule in the predicted molecular composition until the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set.
  • the method preferably further includes:
  • any of the input flow does not meet the preset input flow range of the corresponding petroleum processing device, adjust the preset raw material ratio, and re-use the corresponding fractions as petroleum according to the adjusted preset raw material ratio.
  • the processing raw materials are respectively input to the product prediction model of the corresponding petroleum processing device; until each of the input flow rates meets the preset input flow range of the corresponding petroleum processing device.
  • each input flow rate meets the preset input flow rate range of the corresponding petroleum processing device, it is considered that the subsequent steps can be performed to obtain the predicted molecular composition and the predicted molecular composition of the corresponding predicted product.
  • the steps for predicting the molecular content of each single molecule in the molecular composition are described.
  • the judging whether the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set preferably includes:
  • the steps of obtaining the predicted molecular composition of the corresponding predicted product and the predicted molecular content of each single molecule in the predicted molecular composition are executed.
  • the method preferably further includes:
  • Each of the predicted products is used as a product blending raw material for blending according to a set of preset rules to obtain the molecular composition of multiple sets of mixed products and the content of each single molecule in the mixed product;
  • the product physical properties of each group of mixed products are calculated according to the molecular composition of each group of mixed products and the content of each single molecule in the mixed products.
  • the judging whether the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set preferably includes:
  • the target parameter meets the preset condition, determine that the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set, and output the preset raw material ratio and product prediction model Set with preset rules as the production and processing plan;
  • the operating parameters in the product prediction model and the preset rules in the preset rule set are adjusted, and multiple sets of mixed products are re-obtained until each set of mixed products is The product properties meet the preset product properties, and all the target parameters in the mixed product meet the preset conditions.
  • said obtaining target parameters according to all said mixed products and determining whether said target parameters meet a preset condition preferably includes:
  • the operating parameters include the temperature of the environment in which the reaction path in the product prediction model is located, and the adjustment of the operating parameters in the product prediction model preferably includes:
  • the operating parameter includes the pressure of the environment in which the reaction path in the product prediction model is located, and the adjustment of the operating parameter in the product prediction model preferably includes:
  • the calculation of the product physical properties of each group of the mixed product according to the molecular composition of each group of the mixed product and the content of each single molecule respectively preferably includes:
  • the first component of each group of mixed products is obtained Two-molecule composition and the content of the second component of each single molecule in each group of the product blending raw materials;
  • the calculation of the physical properties of each single molecule preferably includes:
  • the number of groups of each group constituting the single molecule and the contribution value of each group to the physical properties are input into a pre-trained physical property calculation model to obtain the physical properties of the single molecule output by the physical property calculation model ;in,
  • the physical property calculation model is used to calculate the physical properties of the single molecule according to the number of groups of each group contained in the single molecule and the contribution value of each group to the physical properties.
  • the method preferably further includes:
  • the group quantity of each group constituting the single molecule is compared with the molecular information of the template single molecule with known physical properties pre-stored in the database; the molecular information includes: each type of the template single molecule constituting the template single molecule The number of groups;
  • the template single molecule is the same as the single molecule, output the physical properties of the template single molecule as the physical properties of the single molecule;
  • the template single molecule that is the same as the single molecule does not exist, perform the number of groups of each group that will constitute the single molecule and the contribution value of each of the groups to the physical properties, and enter it in advance. The steps of training the physical property calculation model.
  • the step of training the physical property calculation model preferably includes:
  • the physical property calculation model If the deviation value between the predicted physical property and the known physical property is less than the preset deviation threshold, it is determined that the physical property calculation model has converged, and each group pair is obtained from the converged physical property calculation model. The contribution value of the physical property, and the contribution value of the group to the physical property is stored;
  • the contribution value of each group in the physical property calculation model to the physical property is adjusted until the physical property calculation model Convergence.
  • said obtaining the group quantity of each group constituting the sample single molecule preferably includes:
  • a plurality of groups that exist at the same time and contribute to the same physical property are regarded as a multi-level group, and the number of the plurality of groups is regarded as the level of the multi-level group.
  • the physical property calculation model preferably determines the physical properties of a single molecule in the following manner:
  • the physical properties of the single molecule are obtained according to the sum of the corresponding products of various groups.
  • f is the physical property of the single molecule
  • n i is the number of groups of the i-th group in the single molecule
  • ⁇ f i is the contribution value of the i-th group in the single molecule to the physical property
  • a is the correlation constant
  • the obtaining the number of groups of each group constituting the single molecule includes:
  • a plurality of groups that exist at the same time and contribute to the same physical property are regarded as a multi-level group, and the number of the plurality of groups is regarded as the level of the multi-level group.
  • the physical property calculation model determines the physical properties of single molecules in the following manner:
  • each level of groups the product of the number of groups contained in each group and the contribution of each group to the physical properties is obtained, and then the sum of the corresponding products of the various groups is obtained as Contribution value of the group at this level to physical properties;
  • the physical properties of the single molecule are obtained according to the sum of the contribution values of the various groups of the physical properties.
  • the physical property calculation model is as follows:
  • f is the physical properties of the single molecule
  • m 1i is the number of groups of the i-th group in the primary group
  • ⁇ f 1i is the contribution value of the i-th group in the primary group to the physical properties
  • m 2j Is the number of groups of the jth group in the secondary group
  • ⁇ f 2j is the contribution value of the jth group in the secondary group to the physical properties
  • m Nl is the group of the lth group in the N-level group The number of groups, ⁇ f Nl is the contribution value of the first group in the N-level group to the physical properties
  • a is the correlation constant
  • N is a positive integer greater than or equal to 2.
  • the physical properties of the single molecule preferably include: the boiling point of the single molecule;
  • the calculation of the physical properties of the single molecule includes:
  • T is the boiling point of the single molecule
  • SOL is a single molecule vector converted according to the number of groups of each group constituting the single molecule
  • GROUP 11 is the value based on the contribution of the primary group to the boiling point
  • the converted first contribution value vector GROUP 12 is the second contribution value vector converted according to the contribution value of the secondary group to the boiling point
  • GROUP 1N is the second contribution value vector converted according to the contribution value of the N-level group to the boiling point N contribution value vector
  • Numh is the number of atoms in a single molecule other than hydrogen atoms
  • d is the first preset constant
  • b is the second preset constant
  • c is the third preset constant
  • the N is greater than or equal to A positive integer of 2.
  • the physical properties of the single molecule preferably include: the density of the single molecule;
  • the physical property calculation model preferably determines the density of the single molecule in the following manner:
  • a single molecule vector converted according to the number of groups of each group constituting the single molecule
  • the density of the single molecule is obtained according to the proportion of the product of the single molecule vector and the contribution value vector of the primary group in the sum of the corresponding products of the single molecule vector and each level of the group.
  • the density of the single molecule is calculated according to the following physical property calculation model:
  • D is the density of the single molecule
  • SOL is the single molecule vector converted according to the number of groups of each group constituting the single molecule
  • GROUP 21 is the value according to the contribution of the primary group to the density
  • the transformed N+1th contribution value vector GROUP 22 is the N+2th contribution value vector transformed according to the contribution value of the secondary group to the density
  • GROUP 2N is the contribution value of the N-level group to the density In the transformed 2N contribution value vector
  • e is the fourth preset constant; the N is a positive integer greater than or equal to 2.
  • the physical properties of the single molecule preferably include: the octane number of the single molecule;
  • the physical property calculation model preferably determines the octane number of a single molecule in the following manner:
  • a single molecule vector converted according to the number of groups of each group constituting the single molecule
  • a vector of the contribution value of the group of that level converted according to the contribution value of the group of each level to the octane number
  • the octane number of the single molecule is obtained according to the sum of the products of the single molecule vector and the corresponding groups of each level.
  • the octane number of the single molecule is calculated according to the following physical property calculation model:
  • X is the octane number of the single molecule
  • SOL is the single molecule vector converted according to the number of groups of each group constituting the single molecule
  • GROUP 31 is the octane number of the first group
  • GROUP 32 is the 2N+2 contribution value vector converted according to the contribution value of the secondary group to the octane number
  • GROUP 3N is the contribution value vector according to the N level group to the octane
  • the 3Nth contribution value vector obtained by transforming the contribution value of the value; the N is a positive integer greater than or equal to 2; h is the fifth preset constant.
  • the product physical properties of the mixed product preferably include density, cloud point, pour point, aniline point and octane number.
  • calculating the product physical property of each group of the mixed product preferably includes:
  • the density of the group of mixed products is obtained.
  • density is the density of the mixed product
  • D i is the density of the i-th single molecule
  • x i_volume is the second component content of the i-th single molecule
  • calculating the product physical property of each group of the mixed product preferably includes:
  • For each group of mixed products calculate the cloud point contribution value of each of the single molecules in the group of mixed products according to the density and boiling point of each of the single molecules in the group of mixed products;
  • the cloud point of the group of mixed products is calculated.
  • calculating the product physical property of each group of the mixed product preferably includes:
  • For each group of mixed products calculate the pour point contribution value of each of the single molecules in the group of mixed products according to the density and molecular weight of each of the single molecules in the group of mixed products;
  • the pour point of the group of mixed products is calculated.
  • calculating the product physical property of each group of the mixed product preferably includes:
  • the aniline point contribution value of the single molecule is calculated according to the density and boiling point of the single molecule in the group of mixed products;
  • the aniline point of the group of mixed products is calculated.
  • calculating the product physical properties of each group of the mixed products preferably includes:
  • the octane number of the mixed product is calculated by the following formula:
  • the ON is the octane number of the mixed product
  • HISQFG is the molecular collection
  • H is the molecular collection of normal alkanes
  • I is the molecular collection of isoalkanes
  • S is the molecular collection of cycloalkanes
  • Q is the molecular collection of olefins.
  • F is the molecular collection of aromatic hydrocarbons
  • G is the molecular collection of oxygen-containing compounds
  • ⁇ i is the content of each molecule in the mixed product
  • ⁇ H , ⁇ I , ⁇ S , ⁇ Q , ⁇ F , ⁇ G is the total content of normal paraffins, the total content of isoparaffins, the total content of cycloalkanes, the total content of olefins, the total content of aromatic hydrocarbons and the total content of oxygen-containing compounds in the mixed product respectively
  • ⁇ i regression parameters for each molecule of the mixing product ON i is the octane number of each molecule in the product mix
  • C H represents the coefficient of normal paraffins to interact with other molecules
  • C I represents isoparaffin The interaction coefficient with other molecules
  • C S represents the interaction coefficient between cycloalkanes and other molecules
  • C Q represents the interaction coefficient between olefins and other molecules
  • C F represents the interaction coefficient between aromatic hydro
  • the interaction coefficient of other molecules Represents the first constant coefficient between normal paraffin and isoparaffin, Represents the first constant coefficient between normal alkanes and cycloalkanes, Represents the first constant coefficient between normal alkanes and alkenes, Represents the first constant coefficient between normal alkanes and aromatic hydrocarbons, Represents the first constant coefficient between n-alkane and oxygen-containing compound, Represents the first constant coefficient between isoparaffin and cycloalkane, Represents the first constant coefficient between isoparaffin and olefin, Represents the first constant coefficient between isoparaffin and aromatic hydrocarbon, Represents the first constant coefficient between isoparaffin and oxygen-containing compound, Represents the first constant coefficient between cycloalkane and olefin, Represents the first constant coefficient between cycloalkane and aromatic hydrocarbon, Represents the first constant coefficient between cycloalkane and oxygen-containing compound, Represents the first constant coefficient between cycloalkane and oxygen-
  • the step of training the product prediction model preferably includes:
  • the product prediction model includes: a reaction rule set including multiple reaction rules and a reaction rate algorithm;
  • the sample raw material information of the sample raw material preferably includes: the molecular composition of the sample raw material, the molecular content of each molecule in the sample raw material, the molecular composition of the actual product corresponding to the sample raw material, and the actual product The actual content of each molecule in.
  • using the sample raw material information to train the reaction rule set preferably includes:
  • the device output product includes: Describe the sample raw materials, intermediate products and predicted products
  • calculating the first relative deviation according to the first molecular composition of the output product of the device and the second molecular composition of the actual product preferably includes:
  • the second set is not a subset of the first set, acquiring a pre-stored relative deviation value that does not meet a preset condition as the first relative deviation value;
  • the first relative deviation is calculated by the following method: according to the number of types of single molecules in the molecular composition of the predicted product that are not in the second set The proportion of the total number of single molecules in the molecular composition of the predicted product determines the first relative deviation.
  • the first relative deviation is calculated by the following calculation formula:
  • x 1 is the first relative deviation
  • M is the first set
  • M 1 is the set of single molecules in the molecular composition of the sample material
  • M 2 is the single molecule in the molecular composition of the intermediate product consisting of a set of species
  • M 3 of the second set, card represents the number of elements in a set.
  • using the sample raw material information to train the reaction rate algorithm preferably includes:
  • reaction rate algorithm respectively calculate the reaction rate of the reaction path corresponding to each molecule in the molecular composition of the sample raw material
  • respectively calculating the reaction rate of the reaction path corresponding to each molecule in the molecular composition of the sample raw material preferably includes:
  • reaction rate constant is determined based on the calculation method of transition state theory.
  • reaction rate constant is determined according to the following calculation formula:
  • k is the reaction rate constant
  • k B is the Boltzmann constant
  • h is the Planck constant
  • R is the ideal gas constant
  • E is the temperature value of the environment in which the reaction path is located
  • exp is the natural constant as the base
  • ⁇ S is the entropy change before and after the reaction corresponding to the reaction rule corresponding to the reaction path
  • ⁇ E is the reaction energy barrier corresponding to the reaction rule corresponding to the reaction path
  • P is the pressure value of the environment where the reaction path is located
  • is the pressure influence factor corresponding to the reaction rule corresponding to the reaction path.
  • the molecular composition of the different fractions obtained by distillation of the obtained crude oil preferably includes:
  • the types of the petroleum processing equipment include:
  • Catalytic cracking unit delayed coking unit, residual oil hydrogenation unit, hydrocracking unit, diesel hydro-upgrading unit, diesel hydro-refining unit, gasoline hydro-refining unit, catalytic reforming unit and alkylation unit; among them,
  • Each petroleum processing device corresponds to a set of reaction rules.
  • an embodiment of the present invention provides a real-time optimization device for a molecular-level device, and the real-time optimization device includes:
  • the first obtaining unit is used to obtain the molecular composition of crude oil
  • the first processing unit is configured to obtain the molecular composition of different fractions obtained by distillation of the crude oil according to the physical properties of various single molecules in the molecular composition of the crude oil;
  • the second processing unit is used to use the respective fractions as the raw material for petroleum processing according to the preset raw material ratio, and respectively input the pre-trained product prediction model corresponding to the petroleum processing device to obtain the corresponding prediction output by the product prediction model
  • the predicted molecular composition of the product and the predicted molecular content of each single molecule in the predicted molecular composition are used to use the respective fractions as the raw material for petroleum processing according to the preset raw material ratio, and respectively input the pre-trained product prediction model corresponding to the petroleum processing device to obtain the corresponding prediction output by the product prediction model
  • the second obtaining unit is used to obtain a preset standard set of a preset target product
  • the third processing unit based on the predicted molecular composition of the predicted product and the predicted molecular content of each single molecule in the predicted molecular composition, to determine whether the predicted product meets the target corresponding to the predicted product in the preset standard set
  • the preset standard of the product if the predicted product does not meet any of the preset standards of the target product corresponding to the predicted product in the preset standard set, adjust the operating parameters in the product prediction model to renew Obtain the predicted molecular composition of the predicted product and the predicted molecular content of each single molecule in the predicted molecular composition, until the predicted product meets the preset of the target product corresponding to the predicted product in the preset standard set standard.
  • the device further includes:
  • the flow control unit is used to obtain the input flow of the petroleum processing raw materials input to each of the petroleum processing devices; determine whether each of the input flows meets the preset input flow range of the corresponding petroleum processing device; if any If the input flow rate does not meet the preset input flow rate range of the corresponding petroleum processing device, the preset raw material ratio is adjusted, and the corresponding fractions are re-input as petroleum processing raw materials according to the adjusted preset raw material ratio.
  • the product prediction model of the corresponding petroleum processing device until each input flow rate meets the preset input flow range of the corresponding petroleum processing device.
  • the third processing unit is specifically configured to calculate the physical properties of each single molecule in the predicted molecular composition according to the predicted molecular composition of the predicted product and the predicted molecular content of each single molecule in the predicted molecular composition; The predicted physical properties of each single molecule in the predicted molecular composition and the predicted molecular content of each single molecule in the predicted molecular composition, calculate the predicted physical properties of the predicted product; determine whether the predicted physical properties of each predicted product meet the preset The preset physical property limit interval of the corresponding target product in the standard set.
  • the device further includes:
  • the product blending unit is used to blend each of the predicted products as product blending raw materials according to a set of preset rules to obtain the molecular composition of multiple groups of mixed products and the content of each single molecule in the mixed product; According to the molecular composition of the mixed product and the content of each single molecule in the mixed product, the product physical properties of each group of the mixed product are calculated respectively.
  • the third processing unit is specifically configured to determine whether the product physical properties of each group of the mixed products meet the preset product physical properties of the target mixed product obtained by blending the corresponding target products in the preset standard set;
  • the target parameter meets the preset condition, determine that the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set, and output the preset raw material ratio and product prediction model Set with preset rules as the production and processing plan;
  • the operating parameters in the product prediction model and the preset rules in the preset rule set are adjusted, and multiple sets of mixed products are re-obtained until each set of mixed products is The product properties meet the preset product properties, and all the target parameters in the mixed product meet the preset conditions.
  • the third processing unit is specifically used to obtain the product price of each group of mixed products and the output of each group of mixed products; calculate the price of each group of mixed products according to the output of each group of mixed products and the product price of each group of mixed products Product benefit; accumulate the product benefits of each group of mixed products to obtain a cumulative benefit; obtain the raw material price of each group of the petroleum processing raw materials and the operating cost of each petroleum processing device; subtract the cumulative benefit from all petroleum processing
  • the raw material prices of the raw materials and the operating costs of all petroleum processing equipment obtain comprehensive benefits; use the comprehensive benefits as the target parameter; determine whether the comprehensive benefits reach the maximum; if the comprehensive benefits reach the maximum, then It is determined that the target parameter meets the preset condition; if the comprehensive benefit does not reach the maximum value, it is determined that the target parameter does not meet the preset condition.
  • the third processing unit is specifically configured to adjust the temperature of the environment in which the reaction path corresponding to the predicted product in the product prediction model is located; reacquire the predicted molecular composition and the predicted molecular composition of the predicted product according to the adjusted temperature The predicted molecule content of each single molecule in each group of the predicted products until the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set.
  • the third processing unit is specifically configured to adjust the pressure of the environment in which the reaction path corresponding to the predicted product in the product prediction model is located; re-acquire the predicted molecular composition and the predicted molecular composition of the predicted product according to the adjusted pressure The predicted molecular content of each single molecule in the predicted product until the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set.
  • the product blending unit is specifically used to obtain the first molecular composition of each group of the product blending raw materials and the first component content of each single molecule in each group of the product blending raw materials;
  • the product blending unit is specifically used for obtaining the number of groups of each group constituting the single molecule for each single molecule, and obtaining the contribution value of each group to the physical properties;
  • the number of groups of each group of the single molecule and the contribution value of each group to the physical properties are input to a pre-trained physical property calculation model to obtain the physical properties of the single molecule output by the physical property calculation model; wherein
  • the physical property calculation model is used to calculate the physical properties of the single molecule based on the number of groups of each group contained in the single molecule and the contribution value of each group to the physical properties.
  • the device further includes:
  • the single-molecule physical property template matching unit is used to compare the number of groups of each group constituting the single-molecule with the molecular information of template single-molecules with known physical properties pre-stored in the database; the molecular information includes: The number of groups of each group constituting the template single molecule; judge whether there is the same template single molecule as the single molecule; if there is the same template single molecule as the single molecule, output the The physical properties of the template single molecule are taken as the physical properties of the single molecule; if the template single molecule that is the same as the single molecule does not exist, the product blending unit is used to perform the process that will constitute the single molecule.
  • the number of groups and the contribution value of each of the groups to the physical properties are input to the steps of the pre-trained physical properties calculation model.
  • the device further includes:
  • the model training unit is used to construct a single-molecule physical property calculation model; to obtain the number of groups of each group constituting the sample single molecule; wherein the physical properties of the sample single molecule are known; The number of groups of each group is input into the physical property calculation model; the predicted physical property of the sample single molecule output by the physical property calculation model is obtained; if the deviation between the predicted physical property and the known physical property is less than the predicted physical property If a deviation threshold is set, it is determined that the physical property calculation model has converged, the contribution value of each group to the physical property is obtained in the converged physical property calculation model, and the contribution value of the group to the physical property is stored; If the deviation value between the predicted physical property and the known physical property is greater than or equal to the deviation threshold, the contribution value of each group in the physical property calculation model to the physical property is adjusted until the physical property calculation model Convergence.
  • f is the physical property of the single molecule
  • n i is the number of groups of the i-th group in the single molecule
  • ⁇ f i is the contribution value of the i-th group in the single molecule to the physical property
  • a is the correlation constant
  • the model training unit is specifically used to determine the number of primary groups, the number of primary groups, the number of multilevel groups, and the number of groups of multilevel groups among all the groups of the single molecule; All groups constituting a single molecule are regarded as primary groups; multiple groups that exist at the same time and contribute to the same physical property are regarded as multilevel groups, and the number of the multiple groups is regarded as the multilevel group. The level of the group.
  • model training unit is specifically used to establish the following physical property calculation model:
  • f is the physical properties of the single molecule
  • m 1i is the number of groups of the i-th group in the primary group
  • ⁇ f 1i is the contribution value of the i-th group in the primary group to the physical properties
  • m 2j Is the number of groups of the jth group in the secondary group
  • ⁇ f 2j is the contribution value of the jth group in the secondary group to the physical properties
  • m Nl is the group of the lth group in the N-level group The number of groups, ⁇ f Nl is the contribution value of the first group in the N-level group to the physical properties
  • a is the correlation constant
  • N is a positive integer greater than or equal to 2.
  • the product blending unit is specifically used in all groups of the single molecule to determine the number of primary groups, the number of primary groups, the number of multilevel groups, and the number of groups of multilevel groups; will constitute All groups of a single molecule are regarded as primary groups; multiple groups that exist at the same time and contribute to the same physical property are regarded as multilevel groups, and the number of the multiple groups is regarded as the multilevel groups Level.
  • the product blending unit is specifically used to calculate the boiling point of the single molecule according to the following physical property calculation model:
  • T is the boiling point of the single molecule
  • SOL is a single molecule vector converted according to the number of groups of each group constituting the single molecule
  • GROUP 11 is the value based on the contribution of the primary group to the boiling point
  • the converted first contribution value vector GROUP 12 is the second contribution value vector converted according to the contribution value of the secondary group to the boiling point
  • GROUP 1N is the second contribution value vector converted according to the contribution value of the N-level group to the boiling point N contribution value vector
  • Numh is the number of atoms in a single molecule other than hydrogen atoms
  • d is the first preset constant
  • b is the second preset constant
  • c is the third preset constant
  • the N is greater than or equal to A positive integer of 2.
  • the product blending unit is specifically used to calculate the density of the single molecule according to the following physical property calculation model:
  • D is the density of the single molecule
  • SOL is the single molecule vector converted according to the number of groups of each group constituting the single molecule
  • GROUP 21 is the value according to the contribution of the primary group to the density
  • the transformed N+1th contribution value vector GROUP 22 is the N+2th contribution value vector transformed according to the contribution value of the secondary group to the density
  • GROUP 2N is the contribution value of the N-level group to the density In the transformed 2N contribution value vector
  • e is the fourth preset constant; the N is a positive integer greater than or equal to 2.
  • X is the octane number of the single molecule
  • SOL is the single molecule vector converted according to the number of groups of each group constituting the single molecule
  • GROUP 31 is the octane number of the first group
  • GROUP 32 is the 2N+2 contribution value vector converted according to the contribution value of the secondary group to the octane number
  • GROUP 3N is the contribution value vector according to the N level group to the octane
  • the 3Nth contribution value vector obtained by transforming the contribution value of the value; the N is a positive integer greater than or equal to 2; h is the fifth preset constant.
  • the product physical properties of the mixed product include density, cloud point, pour point, aniline point and octane number.
  • density is the density of the mixed product
  • D i is the density of the i-th single molecule
  • x i_volume is the second component content of the i-th single molecule
  • the product blending unit is specifically used to calculate the cloud point contribution value of each single molecule according to the density and boiling point of each single molecule in the group of mixed products for each group of mixed products; Set the cloud point contribution value of all the single molecules in the mixed product and the content of each single molecule to calculate the cloud point of the mixed product.
  • the product blending unit is specifically configured to calculate the pour point contribution value of each single molecule according to the density and molecular weight of each single molecule in the group of mixed products for each group of mixed products; Group the pour point contribution value of all the single molecules in the mixed product and the content of each single molecule to calculate the pour point of the mixed product.
  • the product blending unit is specifically used to calculate the aniline point contribution value of the single molecule according to the density and boiling point of the single molecule in the group of mixed products for each group of mixed products;
  • the aniline point contribution value of all the single molecules and the content of each single molecule are calculated to calculate the aniline point of the mixed product.
  • the product blending unit is specifically used to obtain the octane number of each single molecule and the content of each single molecule in the group of mixed products for each group of mixed products; the mixed product is calculated by the following formula The octane rating:
  • the ON is the octane number of the mixed product
  • HISQFG is the molecular collection
  • H is the molecular collection of normal alkanes
  • I is the molecular collection of isoalkanes
  • S is the molecular collection of cycloalkanes
  • Q is the molecular collection of olefins.
  • F is the molecular collection of aromatic hydrocarbons
  • G is the molecular collection of oxygen-containing compounds
  • ⁇ i is the content of each molecule in the mixed product
  • ⁇ H , ⁇ I , ⁇ S , ⁇ Q , ⁇ F , ⁇ G is the total content of normal paraffins, the total content of isoparaffins, the total content of cycloalkanes, the total content of olefins, the total content of aromatic hydrocarbons and the total content of oxygen-containing compounds in the mixed product respectively
  • ⁇ i regression parameters for each molecule of the mixing product ON i is the octane number of each molecule in the product mix
  • C H represents the coefficient of normal paraffins to interact with other molecules
  • C I represents isoparaffin The interaction coefficient with other molecules
  • C S represents the interaction coefficient between cycloalkanes and other molecules
  • C Q represents the interaction coefficient between olefins and other molecules
  • C F represents the interaction coefficient between aromatic hydro
  • the interaction coefficient of other molecules Represents the first constant coefficient between normal paraffin and isoparaffin, Represents the first constant coefficient between normal alkanes and cycloalkanes, Represents the first constant coefficient between normal alkanes and alkenes, Represents the first constant coefficient between normal alkanes and aromatic hydrocarbons, Represents the first constant coefficient between n-alkane and oxygen-containing compound, Represents the first constant coefficient between isoparaffin and cycloalkane, Represents the first constant coefficient between isoparaffin and olefin, Represents the first constant coefficient between isoparaffin and aromatic hydrocarbon, Represents the first constant coefficient between isoparaffin and oxygen-containing compound, Represents the first constant coefficient between cycloalkane and olefin, Represents the first constant coefficient between cycloalkane and aromatic hydrocarbon, Represents the first constant coefficient between cycloalkane and oxygen-containing compound, Represents the first constant coefficient between cycloalkane and oxygen-
  • the device further includes:
  • the model training unit is used to establish a product prediction model; wherein the product prediction model includes: a reaction rule set including a variety of reaction rules and a reaction rate algorithm; acquiring sample raw material information of the sample raw material; using the sample raw material information, Train the reaction rule set, and fix the reaction rule set that has been trained; use the sample material information to train the reaction rate algorithm, and fix the reaction rate algorithm that has been trained to obtain the training completion The product prediction model.
  • the product prediction model includes: a reaction rule set including a variety of reaction rules and a reaction rate algorithm; acquiring sample raw material information of the sample raw material; using the sample raw material information, Train the reaction rule set, and fix the reaction rule set that has been trained; use the sample material information to train the reaction rate algorithm, and fix the reaction rate algorithm that has been trained to obtain the training completion The product prediction model.
  • the sample raw material information of the sample raw material includes: the molecular composition of the sample raw material, the molecular content of each molecule in the sample raw material, the molecular composition of the actual product corresponding to the sample raw material, and the molecular composition of the actual product The actual content of each molecule.
  • the model training unit is specifically configured to process the molecular composition of the sample raw material according to a preset reaction rule set to obtain the reaction path corresponding to each molecule in the molecular composition of the sample raw material;
  • the reaction path corresponding to each molecule in the molecular composition of the raw material obtains the first molecular composition of the device output product including the sample raw material, the intermediate product, and the predicted product;
  • the device output product includes: the sample raw material, Intermediate products and predicted products; calculate the first relative deviation based on the first molecular composition of the output product of the device and the second molecular composition of the actual product; if the first relative deviation meets a preset condition, fix the Reaction rule set; if the first relative deviation does not meet the preset condition, adjust the reaction rule in the reaction rule set, and recalculate the first relative difference according to the adjusted reaction rule set until the The first relative deviation meets the preset condition.
  • the model training unit is specifically used to obtain the types of single molecules in the first molecular composition to form a first set; to obtain the types of single molecules in the second molecular composition to form a second set; to determine the Whether the second set is a subset of the first set; if the second set is not a subset of the first set, obtain a pre-stored relative deviation value that does not meet the preset condition as the first relative Deviation value; if the second set is a subset of the first set, the first relative deviation is calculated by the following calculation formula:
  • x 1 is the first relative deviation
  • M is the first set
  • M 1 is the set of single molecules in the molecular composition of the sample material
  • M 2 is the single molecule in the molecular composition of the intermediate product consisting of a set of species
  • M 3 of the second set, card represents the number of elements in a set.
  • the model training unit is specifically configured to calculate the reaction rate of the reaction path corresponding to each molecule in the molecular composition of the sample material according to the reaction rate algorithm; according to the molecule of each molecule in the sample material The content and the reaction rate corresponding to the reaction path of the molecule are used to obtain the predicted content of each molecule in the predicted product corresponding to the sample raw material; according to the predicted content of each molecule in the predicted product and the predicted content of each molecule in the actual product The actual content of the molecule, the second relative deviation is calculated; if the second relative deviation meets the preset condition, the reaction rate algorithm is fixed; if the second relative deviation does not meet the preset condition, the reaction rate is adjusted For the parameters in the algorithm, the second relative deviation is recalculated according to the adjusted reaction rate algorithm until the second relative deviation meets the preset condition.
  • model training unit is specifically configured to calculate the reaction rate of each reaction path according to the reaction rate constant in the reaction rate algorithm
  • reaction rate constant is determined according to the following calculation formula:
  • k is the reaction rate constant
  • k B is the Boltzmann constant
  • h is the Planck constant
  • R is the ideal gas constant
  • E is the temperature value of the environment in which the reaction path is located
  • exp is the natural constant as the base
  • ⁇ S is the entropy change before and after the reaction corresponding to the reaction rule corresponding to the reaction path
  • ⁇ E is the reaction energy barrier corresponding to the reaction rule corresponding to the reaction path
  • P is the pressure value of the environment where the reaction path is located
  • is the pressure influence factor corresponding to the reaction rule corresponding to the reaction path.
  • the types of petroleum processing equipment include: catalytic cracking equipment, delayed coking equipment, residual oil hydrogenation equipment, hydrocracking equipment, diesel hydro-upgrading equipment, diesel hydro-refining equipment, gasoline hydro-refining equipment, and catalytic cracking equipment. Reforming unit and alkylation unit; among them, each type of petroleum processing unit corresponds to a set of reaction rules.
  • the present invention provides a real-time optimization system for molecular-level devices.
  • the real-time optimization system for molecular-level devices includes a processor and a memory; the processor is used to perform real-time optimization of molecular-level devices stored in the memory.
  • the optimization program is used to realize the real-time optimization method of the molecular-level device described in the first aspect.
  • the present invention provides a computer-readable storage medium that stores one or more programs, and the one or more programs can be executed by one or more processors to realize The method for real-time optimization of molecular-level devices described in the first aspect.
  • the method provided by the embodiment of the present invention obtains the molecular composition of crude oil; obtains the molecular composition of different fractions obtained by distillation of the crude oil according to the physical properties of various single molecules in the molecular composition of the crude oil; according to a preset raw material ratio , Using the respective fractions as raw materials for petroleum processing, respectively input into the pre-trained product prediction model corresponding to the petroleum processing device to obtain the predicted molecular composition and the predicted molecular composition of the corresponding predicted product output by the product prediction model The predicted molecular content of each single molecule; obtaining a preset standard set of preset target products; judging whether the predicted product is based on the predicted molecular composition of the predicted product and the predicted molecular content of each single molecule in the predicted molecular composition Meet the preset standard of the target product corresponding to the predicted product in the preset standard set; if the predicted product does not meet any preset standard of the target product corresponding to the predicted product in the preset standard set Standard, the operating parameters in the product prediction model
  • FIG. 1 is a schematic flowchart of a method for real-time optimization of a molecular-level device according to an embodiment of the present invention.
  • Fig. 2 is a schematic structural diagram of a real-time optimization device for a molecular-level device according to an embodiment of the present invention.
  • FIG. 3 is a structural diagram of a real-time optimization system for a molecular-level device according to another embodiment of the present invention.
  • the embodiment of the present invention provides a method for real-time optimization of a molecular-level device. As shown in FIG. 1, the method may include the following steps:
  • the raw material molecular composition of the petroleum processing raw material that is, the information of various molecules (single molecules) included in the petroleum processing raw material.
  • the raw material molecular composition of the petroleum processing raw material is a molecular composition based on SOL.
  • the types of single molecules include, but are not limited to: alkenes, alkanes, cycloalkanes, and aromatic hydrocarbons.
  • the types of petroleum processing equipment include:
  • Catalytic cracking unit delayed coking unit, residual oil hydrogenation unit, hydrocracking unit, diesel hydro-upgrading unit, diesel hydro-refining unit, gasoline hydro-refining unit, catalytic reforming unit and alkylation unit; among them,
  • Each petroleum processing device corresponds to a set of reaction rules.
  • the preset standard set includes one or more preset standards, where the preset standards include, but are not limited to: the comprehensive benefits of the generated products, and the proportion of the generated amount of the predicted products in the mixed product , The predicted physical properties corresponding to the predicted product.
  • the preset standards include, but are not limited to: the comprehensive benefits of the generated products, and the proportion of the generated amount of the predicted products in the mixed product , The predicted physical properties corresponding to the predicted product.
  • the different preset standards will be described later and will not be repeated here.
  • step S104 According to the predicted molecular composition of the predicted product and the predicted molecular content of each single molecule, determine whether the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set. If it meets, step S105 is executed; if the predicted product does not meet any preset standard of the target product corresponding to the predicted product in the preset standard set, step S106 is executed.
  • the operating parameters in the embodiment of the present invention include the temperature of the environment where the reaction path is located in the product prediction model, and the pressure of the environment where the reaction path is located in the product prediction model. The adjustment of the operating parameters will be described later, and will not be repeated here.
  • the method further includes:
  • the respective fraction products are used as the petroleum processing raw material according to the preset raw material ratio.
  • the present invention can be detected by comprehensive two-dimensional gas chromatography, quadrupole gas chromatography-mass spectrometry, gas chromatography/field ionization-time-of-flight mass spectrometry, gas chromatography, near-infrared spectroscopy, nuclear magnetic resonance
  • spectroscopy Raman spectroscopy, Fourier transform ion cyclotron resonance mass spectrometry, electrostatic field orbitrap mass spectrometry, and ion mobility mass spectrometry.
  • other methods can also be used to determine the molecular composition of petroleum processing raw materials, such as ASTM D2425, SH/T 0606, and ASTM D8144-18.
  • the above-mentioned molecular detection method can detect the structure of the molecule, and thus obtain the type of the molecule.
  • the structure-oriented lumped molecular characterization method is the SOL molecular characterization method, which uses 24 structural incremental fragments to characterize the basis of complex hydrocarbon molecules. structure. Any petroleum molecule can be expressed by a set of specific structural incremental fragments.
  • the SOL method is lumped on the molecular scale, reducing the number of molecules in the actual system from millions to thousands, greatly reducing the complexity of the simulation.
  • This characterization method can not only express alkanes, cycloalkanes, up to complex aromatic structures containing 50-60 carbon atoms, but also express olefins or cycloalkenes as intermediate products or secondary reaction products. In addition, it also considers sulfur, nitrogen, Heteroatom compounds such as oxygen.
  • the molecular composition of crude oil is the information of various molecules (single molecules) in the crude oil.
  • the single molecules contained in the raw materials the types of single molecules, the volume and content of each single molecule, etc.
  • the boiling point of each single molecule in the crude oil can be calculated separately, and the fraction distillation range can be determined based on the boiling point and content of each single molecule, and the crude oil can be distilled and cut according to the fraction distillation range to obtain multiple sets of fractions.
  • the fraction distillation range can be determined based on the boiling point and content of each single molecule, and the crude oil can be distilled and cut according to the fraction distillation range to obtain multiple sets of fractions.
  • the molecular composition of each group of fractions obtained after crude oil distillation can be known.
  • the corresponding fractions are used as petroleum processing raw materials for secondary processing, where the preset raw material ratio is the proportion of each fraction input into different petroleum processing devices, and the product prediction model of each petroleum processing device is used. Combining the molecular composition of the fraction input to the petroleum processing unit, the molecular composition in the predicted product and the content of each single molecule in the predicted product are obtained.
  • the distillate obtained by distillation of crude oil includes light oil and heavy oil.
  • Light oil such as naphtha does not require secondary processing, while heavy oil Generally, different secondary processing is required to convert heavy oil products into light oil products to improve the properties of the oil products.
  • the corresponding fractions are input to the petroleum processing unit according to the preset raw material ratio
  • the preset raw material ratio includes: the type and amount of the distillate input to the petroleum processing device, and the fraction that does not require the secondary processing device is no longer in the preset raw material ratio.
  • the product prediction model has been trained and optimized.
  • the product prediction model can be used to obtain the petroleum processing raw materials after being input to the petroleum processing device, and adjust the reaction conditions in the petroleum processing device, such as conditions such as pressure, temperature, and space velocity.
  • the reaction conditions in the petroleum processing device such as conditions such as pressure, temperature, and space velocity.
  • the product situation under certain set conditions can be obtained.
  • the method further includes:
  • any of the input flow does not meet the preset input flow range of the corresponding petroleum processing device, adjust the preset raw material ratio, and re-use the corresponding fractions as petroleum according to the adjusted preset raw material ratio.
  • the processing raw materials are respectively input into the product prediction model of the corresponding petroleum processing device; until each of the input flows meets the preset input flow range of the corresponding petroleum processing device;
  • each input flow rate meets the preset input flow rate range of the corresponding petroleum processing device, execute the predicted molecular composition of the corresponding predicted product and the predicted molecular content of each single molecule in the predicted molecular composition A step of.
  • the subsequent steps of the scheme are directly performed.
  • each group of petroleum processing equipment has a corresponding processing capacity.
  • the processing time of the raw materials in the petroleum processing equipment is too short and cannot be fully reacted. Bad conditions may cause damage to the petroleum processing device.
  • a preset input flow range is set. The maximum value of the range can be 80 to 9 percent of the maximum processing capacity of the petroleum processing device. During the fifteenth period, by limiting the amount of raw materials entering the petroleum processing equipment, avoiding damage to the petroleum processing equipment.
  • the preset raw material ratio is adjusted, and the amount of petroleum processing raw materials input to the petroleum processing device is re-planned, so that the raw material of each petroleum processing device The input flow rate is in line with the preset input flow range of the corresponding petroleum processing device.
  • the judging whether the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set further includes:
  • the predicted physical properties of the predicted product are calculated; the predicted physical properties of the predicted product include, but are not limited to: density, cloud point, pour point , Aniline point and octane number.
  • the predicted physical property of each predicted product meets the preset physical property limit interval, it is determined that the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set, then execute The step of obtaining the predicted molecular composition of the corresponding predicted product and the predicted molecular content of each single molecule in the predicted molecular composition.
  • the method further includes:
  • Each of the predicted products is used as a product blending raw material for blending according to a set of preset rules to obtain the molecular composition of multiple sets of mixed products and the content of each single molecule in the mixed product;
  • the product physical properties of each group of mixed products are calculated according to the molecular composition of each group of mixed products and the content of each single molecule.
  • the predicted product input by each petroleum processing device is used as the product blending raw material for blending, wherein each set of preset rules in the preset rule set includes the type and quantity of the predicted product used, and
  • the predicted products output by different petroleum processing devices are mixed to obtain corresponding mixed products, where the mixed products include, but are not limited to, gasoline products such as automotive oil, lubricating oil, hydraulic oil, gear oil, and cutting oil for vehicles.
  • the production planning can be completed by blending the raw materials for the blending of various products, so that each blended product obtained meets the national standards of the corresponding product.
  • the molecular composition of the predicted product and the content of each single molecule in the predicted product combined with a preset rule set, the molecular composition of different mixed products and the content of each single molecule in the mixed product are obtained.
  • the judging whether the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set further includes:
  • the target parameter meets the preset condition, determine that the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set, and output the preset raw material ratio and product prediction model Set with preset rules as the production and processing plan;
  • the operating parameters in the product prediction model and the preset rules in the preset rule set are adjusted, and multiple sets of mixed products are re-obtained until each set of mixed products is The product properties meet the preset product properties, and all the target parameters in the mixed product meet the preset conditions.
  • the product physical properties of each group of mixed products are calculated separately.
  • the physical properties of each single molecule in the mixed product can be calculated by determining the various single molecules contained in each group of mixed products, that is, determining the molecular composition of the mixed product. , And then calculate the physical properties of the blended gasoline product based on the physical properties and content of each single molecule in the blended gasoline.
  • the physical properties of a single molecule include, but are not limited to: density, boiling point, density, and octane number.
  • the physical properties of a single molecule can also include: viscosity, solubility parameters, cetane number, degree of unsaturation, and so on.
  • each mixed product blended at this time is a qualified product.
  • the mixed product obtain the relevant target parameters and confirm whether the target parameters are Meet the preset conditions, where the target parameters can be the economic benefits of the product, the content of substances that are harmful to the environment in the product, and the proportion of products that meet a certain preset standard among all mixed products.
  • the ultimate goal of the refinery's refining is to pursue benefits.
  • a gross profit value can be calculated through the price of each mixed product and the quantity of the mixed product. The gross profit value can be used to confirm whether the final benefit has reached the maximum.
  • Whether the target parameters meet the preset conditions and confirm whether the final benefit reaches the maximum can be calculated by random algorithms.
  • the content of substances that are harmful to the environment in the mixed product will also affect the mixture.
  • the calculated benefit value is large, it cannot be sold on the sales side and cannot be converted into benefit. Therefore, in order to increase the competitiveness of oil products, the content of substances harmful to the environment in the mixed product can be determined.
  • the market will have different demand. For example, the price of gasoline for car No. 98 is higher than the price of gasoline for car No. 95, but the price of car gasoline for car No. 95 Gasoline consumption is greater.
  • the refinery produces a large amount of gasoline for the 98th car, but the market will take longer to digest, resulting in a backlog of gasoline for the 98th car, resulting in more manpower and other aspects.
  • the cost, resulting in the final benefit is not as good as the production of No. 95 motor gasoline, so in this step, the production of mixed products that meet a certain preset standard can be calculated as a proportion of all mixed products to avoid products Backlog.
  • the judging whether the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set further includes:
  • the comprehensive benefit is taken as the target parameter to ensure the production benefit, and it can be judged whether the comprehensive benefit reaches the maximum value through the global optimization algorithm of multi-start random search.
  • the target parameters when the target parameters also meet the corresponding preset conditions, it means that the entire production process has met the various production requirements at this time, and sustainable production can be carried out.
  • different fractions are input in the output plan.
  • the product prediction model used to calculate the molecular composition of the predicted products produced by each petroleum processing device and the content of each single molecule, and the prediction of the output of the petroleum processing device
  • the set of preset rules for product blending is used as a production and processing plan.
  • the production and processing plan is used for production, and real-time optimization of the device is realized at the molecular level.
  • the target parameters do not meet the preset conditions, it means that the economic benefits of the final blended product may not reach the maximum value, or the amount of substances that affect the environment in the blended product exceeds the set value. Or, the proportion of mixed products that meet a certain preset standard in the mixed products does not reach the set value.
  • the operating parameters in the product prediction model and the preset rule set can be adjusted To obtain multiple sets of mixed products in another case, until the product properties of each set of mixed products output in this solution meet the preset product properties, and at the same time, the target parameters in all mixed products meet the preset conditions, that is Complete real-time optimization of molecular-level devices.
  • the operating parameters include the temperature of the environment in which the reaction path in the product prediction model is located.
  • adjusting the operating parameters in the product prediction model further includes:
  • the operating parameters include the pressure of the environment in which the reaction path in the product prediction model is located, and the adjusting the operating parameters in the product prediction model further includes:
  • the product properties of each group of mixed products are calculated according to the molecular composition of each group of mixed products and the content of each single molecule, including:
  • the second molecular composition of each group of mixed products and the first component content of each single molecule are obtained.
  • the second component content in this embodiment, the preset rules in the preset rule set set the type and quantity of the required product blending raw materials, through the molecular composition of the product blending raw materials and each single molecule
  • the content of the first component obtains the second molecular composition of the mixed product and the content of the second component of each single molecule.
  • the physical properties of each single molecule are calculated; in this example, for each single molecule , To obtain the number of groups of each group constituting a single molecule, and obtain the contribution value of each group to the physical properties; the number of groups of each group constituting the single molecule and the contribution value of each group to the physical properties , Input the pre-trained physical property calculation model to obtain the physical properties of the single molecule output by the physical property calculation model.
  • blended gasoline products include: research octane number, motor octane number, Reid vapor pressure, Engler's distillation range, density, benzene volume fraction, aromatic hydrocarbon volume fraction, olefin volume fraction, oxygen content and sulfur content .
  • Calculating the physical properties of a single molecule includes: for each single molecule, obtaining the number of groups of each group constituting the single molecule, and obtaining the contribution value of each of the groups to the physical properties;
  • the number of groups of each group constituting the single molecule and the contribution value of each group to the physical properties are input into a pre-trained physical property calculation model to obtain the physical properties of the single molecule output by the physical property calculation model ;in,
  • the physical property calculation model is used to calculate the physical properties of the single molecule according to the number of groups of each group contained in the single molecule and the contribution value of each group to the physical properties.
  • the Methods before the number of groups of each group constituting the single molecule and the contribution value of each group to the physical properties are input into the pre-trained physical properties calculation model, the Methods also include:
  • the group quantity of each group constituting the single molecule is compared with the molecular information of the template single molecule with known physical properties pre-stored in the database; the molecular information includes: each type of the template single molecule constituting the template single molecule The number of groups;
  • the template single molecule is the same as the single molecule, output the physical properties of the template single molecule as the physical properties of the single molecule;
  • the template single molecule that is the same as the single molecule does not exist, perform the number of groups of each group that will constitute the single molecule and the contribution value of each of the groups to the physical properties, and enter it in advance. The steps of training the physical property calculation model.
  • the step of training the physical property calculation model includes:
  • the physical property calculation model If the deviation value between the predicted physical property and the known physical property is less than the preset deviation threshold, it is determined that the physical property calculation model has converged, and each group pair is obtained from the converged physical property calculation model. The contribution value of the physical property, and the contribution value of the group to the physical property is stored;
  • the contribution value of each group in the physical property calculation model to the physical property is adjusted until the physical property calculation model Convergence.
  • the physical property calculation model includes: the contribution value of each group to the physical property.
  • the contribution value is an adjustable value, and the contribution value is the initial value during the first training.
  • the physical property calculation model includes: the contribution value of each group to each physical property.
  • a training sample set is preset.
  • the training sample set includes multiple samples of single-molecule information.
  • the sample single molecule information includes, but is not limited to: the number of groups of each group constituting the sample single molecule, and the physical properties of the sample single molecule.
  • the contribution value of each group to each physical property can be obtained in the converged physical property calculation model.
  • the contribution value of the group to each physical property is stored, so that when the physical properties of a single molecule are subsequently calculated, the contribution value of each group in the single molecule to the physical properties that need to be known can be obtained, and
  • the number of groups of each group of the single molecule and the contribution value of each group to the physical properties that need to be known are used as the input of the physical property calculation model.
  • the physical property calculation model is the number of groups of each group of the single molecule
  • the contribution value of each group to the physical property that needs to be known is used as a model parameter (instead of the adjustable contribution value of each group to the physical property in the physical property calculation model), the physical property that needs to be learned is calculated.
  • the predicted physical properties of the sample single molecule output by the physical property calculation model will also be multiple. In this case, calculate the difference between each predicted physical property and the corresponding known physical property. Determine whether the deviation values between all predicted physical properties and the corresponding known physical properties are less than the preset deviation value. If yes, determine that the physical property calculation model has converged. According to the convergent physical property calculation model, you can obtain each The contribution value of each group corresponding to the physical properties, through the above scheme, the contribution value of each group to different physical properties can be obtained.
  • f is the physical property of the single molecule
  • n i is the number of groups of the i-th group in the single molecule
  • ⁇ f i is the contribution value of the i-th group in the single molecule to the physical property
  • a is the correlation constant
  • the primary group and the multi-level group are determined among all the groups of a single molecule; among them, all the groups constituting the single molecule are regarded as primary groups; those that exist simultaneously and contribute to the same physical property.
  • a variety of groups are regarded as multi-level groups, and the number of multiple groups is regarded as the level of multi-level groups.
  • the contribution value will fluctuate to a certain extent.
  • the way we divide the above-mentioned multi-level groups can also be divided by the chemical bond force between the groups according to the preset bond force interval. For different physical properties, different chemical bond forces will have different effects. The specific can be based on the stability of the molecule. The impact of physical properties is divided.
  • the obtaining the number of groups of each group constituting the single molecule of the sample includes:
  • a plurality of groups that exist at the same time and contribute to the same physical property are regarded as a multi-level group, and the number of the plurality of groups is regarded as the level of the multi-level group.
  • Model 2 Based on the divided multi-level groups, the following physical property calculation models can be established:
  • f is the physical properties of a single molecule
  • m 1i is the number of groups of the i-th group in the primary group
  • ⁇ f 1i is the contribution value of the i-th group in the primary group to the physical properties
  • m 2j is two The number of groups in the j-th group in the second-level group
  • ⁇ f 2j is the contribution value of the j-th group in the second-level group to the physical properties
  • m Nl is the number of the first group in the N-level group
  • ⁇ f Nl is the contribution value of the first group in the N-level group to the physical properties
  • a is the correlation constant
  • N is a positive integer greater than or equal to 2.
  • the obtaining the number of groups of each group constituting the single molecule includes:
  • a plurality of groups that exist at the same time and contribute to the same physical property are regarded as a multi-level group, and the number of the plurality of groups is regarded as the level of the multi-level group.
  • physical property calculation models can also be constructed for each physical property according to different types of physical properties.
  • T is the boiling point of a single molecule
  • SOL is a single molecule vector converted according to the number of groups of each group constituting a single molecule
  • GROUP 1 is the first contribution converted according to the contribution value of the first-level group to the boiling point.
  • Value vector GROUP 2 is the second contribution value vector converted according to the contribution value of the secondary group to the boiling point
  • GROUP N is the Nth contribution value vector converted according to the contribution value of the N-level group to the boiling point
  • Numh is the single The number of atoms in the molecule excluding hydrogen atoms
  • d is the first predetermined constant
  • b is the second predetermined constant
  • c is the third predetermined constant
  • N is a positive integer greater than or equal to 2.
  • the single-molecule vector transformed according to the number of groups of each group constituting a single molecule includes: taking the number of types of all groups constituting a single molecule as the dimension of the single-molecule vector; taking the group of each group The quantity is used as the element value of the corresponding dimension in the single molecule vector.
  • the first contribution value vector converted according to the contribution value of each primary group of a single molecule to the boiling point includes: taking the number of primary group types as the dimension of the first contribution value vector; The contribution value of the group to the boiling point is taken as the element value of the corresponding dimension in the first contribution value vector.
  • the second contribution value vector converted according to the contribution value of each secondary group of a single molecule to the boiling point includes: taking the number of the type of secondary group as the dimension of the second contribution value vector; The contribution value of the group to the boiling point is taken as the element value of the corresponding dimension in the second contribution value vector.
  • the Nth contribution value vector converted from the contribution of each N-level group of a single molecule to the boiling point includes: taking the number of N-level groups as the dimension of the Nth contribution value vector; The contribution value of each N-level group to the boiling point is taken as the element value of the corresponding dimension in the Nth contribution value vector.
  • D is the density of a single molecule
  • SOL is a single molecule vector converted according to the number of groups of each group constituting a single molecule
  • GROUP 21 is the N+th converted according to the contribution value of the first-level group to the density.
  • 1 Contribution value vector GROUP 22 is the N+2th contribution value vector transformed according to the contribution value of the secondary group to the density
  • GROUP 2N is the 2Nth contribution value vector transformed according to the N level group’s contribution to the density
  • E is the fourth preset constant
  • N is a positive integer greater than or equal to 2.
  • the single-molecule vector transformed according to the number of groups of each group constituting a single molecule includes: taking the number of types of all groups constituting a single molecule as the dimension of the single-molecule vector; taking the group of each group The quantity is used as the element value of the corresponding dimension in the single molecule vector.
  • the N+1th contribution value vector transformed according to the contribution value of each primary group of a single molecule to the density including: taking the number of primary groups as the dimension of the N+1th contribution value vector; The contribution value of each primary group to the density is taken as the element value of the corresponding dimension in the N+1th contribution value vector.
  • the N+2th contribution value vector obtained by transforming the contribution value of each secondary group of a single molecule to the density respectively includes: taking the number of types of secondary groups as the dimension of the N+2th contribution value vector; The contribution value of each secondary group to the density is taken as the element value of the corresponding dimension in the N+2th contribution value vector.
  • the 2N contribution value vector obtained by transforming the contribution value of each N-level group of a single molecule to the density respectively includes: taking the number of types of N-level groups as the dimension of the 2N contribution value vector; The contribution value of each N-level group to the density is taken as the element value of the corresponding dimension in the 2N-th contribution value vector.
  • X is the octane number of a single molecule
  • SOL is a single molecule vector converted according to the number of groups of each group constituting a single molecule
  • GROUP 31 is converted according to the contribution value of the primary group to the octane number
  • the 2N+1 contribution value vector of the group GROUP 32 is the 2N+2 contribution value vector converted according to the contribution value of the secondary group to the octane number
  • GROUP 3N is the contribution value of the N-level group to the octane number
  • the single-molecule vector transformed according to the number of groups of each group constituting a single molecule includes: taking the number of types of all groups constituting a single molecule as the dimension of the single-molecule vector; taking the group of each group The quantity is used as the element value of the corresponding dimension in the single molecule vector.
  • the 2N+1th contribution value vector transformed according to the contribution value of each primary group of a single molecule to the octane number including: taking the number of primary groups as the dimension of the 2N+1 contribution value vector ; The contribution value of each primary group to the octane number is taken as the element value of the corresponding dimension in the 2N+1 contribution value vector.
  • the 2N+2 contribution value vector converted according to the contribution value of each secondary group of a single molecule to the octane number including: taking the number of secondary groups as the dimension of the 2N+2 contribution value vector ; The contribution value of each secondary group to the octane number is taken as the element value of the corresponding dimension in the 2N+2 contribution value vector.
  • the 3N contribution value vector converted according to the contribution value of each N-level group of a single molecule to the octane number includes: taking the number of N-level groups as the dimension of the 3N contribution value vector ; Take the contribution value of each N-level group to the octane number as the element value of the corresponding dimension in the 3N contribution value vector.
  • the single molecule is used as a template single molecule, and the number of groups and corresponding physical properties of each group constituting the single molecule are stored in the database.
  • Product physical properties of mixed products including: research octane number, motor octane number, Reid vapor pressure, Engler's distillation range, density, benzene volume fraction, aromatic hydrocarbon volume fraction, olefin volume fraction, oxygen content, and sulfur content Fraction.
  • density is the density of the mixed product
  • D i is the density of the i-th single molecule
  • x i_volume the content of the i-th single molecule
  • Method 2 When the physical property of the mixture is the cloud point, calculate the physical property of the mixture, including:
  • the cloud point of the mixture is calculated.
  • Method 3 When the physical properties of the mixture are the pour point, the physical properties of the mixture are calculated, including:
  • Method 4 When the physical property of the mixture is the aniline point, the physical property of the mixture is calculated, including:
  • the ON is the octane number of the mixed product
  • HISQFG is the molecular collection
  • H is the molecular collection of normal alkanes
  • I is the molecular collection of isoalkanes
  • S is the molecular collection of cycloalkanes
  • Q is the molecular collection of olefins.
  • F is the molecular collection of aromatic hydrocarbons
  • G is the molecular collection of oxygen-containing compounds
  • ⁇ i is the content of each molecule in the mixed product
  • ⁇ H , ⁇ I , ⁇ S , ⁇ Q , ⁇ F , ⁇ G is the total content of normal paraffins, the total content of isoparaffins, the total content of cycloalkanes, the total content of olefins, the total content of aromatic hydrocarbons and the total content of oxygen-containing compounds in the mixed product respectively
  • ⁇ i regression parameters for each molecule of the mixing product ON i is the octane number of each molecule in the product mix
  • C H represents the coefficient of normal paraffins to interact with other molecules
  • C I represents isoparaffin The interaction coefficient with other molecules
  • C S represents the interaction coefficient between cycloalkanes and other molecules
  • C Q represents the interaction coefficient between olefins and other molecules
  • C F represents the interaction coefficient between aromatic hydro
  • the interaction coefficient of other molecules Represents the first constant coefficient between normal paraffin and isoparaffin, Represents the first constant coefficient between normal alkanes and cycloalkanes, Represents the first constant coefficient between normal alkanes and alkenes, Represents the first constant coefficient between normal alkanes and aromatic hydrocarbons, Represents the first constant coefficient between n-alkane and oxygen-containing compound, Represents the first constant coefficient between isoparaffin and cycloalkane, Represents the first constant coefficient between isoparaffin and olefin, Represents the first constant coefficient between isoparaffin and aromatic hydrocarbon, Represents the first constant coefficient between isoparaffin and oxygen-containing compound, Represents the first constant coefficient between cycloalkane and olefin, Represents the first constant coefficient between cycloalkane and aromatic hydrocarbon, Represents the first constant coefficient between cycloalkane and oxygen-containing compound, Represents the first constant coefficient between cycloalkane and oxygen-
  • the product prediction model includes: a reaction rule set including multiple reaction rules and a reaction rate algorithm;
  • a product prediction model is established corresponding to the type of petroleum processing device.
  • the product prediction model corresponding to the petroleum processing device includes: a set of reaction rules and a reaction rate algorithm corresponding to the petroleum processing device.
  • the reaction rule set includes: multiple reaction rules corresponding to the petroleum processing device.
  • the sample raw material information of the sample raw material includes: the molecular composition of the sample raw material, the molecular content of each molecule in the sample raw material, the molecular composition of the actual product corresponding to the sample raw material, and the actual content of each molecule in the actual product.
  • the actual product refers to the product obtained after the sample raw material is processed by the petroleum processing device.
  • the molecular composition of the sample material is processed according to a preset reaction rule set to obtain the reaction path corresponding to each molecule in the molecular composition of the sample material; when the reaction path is calculated for the first time, the molecular composition of the sample material is calculated according to the preset
  • the set of reaction rules are set for processing, and the reaction path corresponding to each molecule in the molecular composition of the sample material is obtained.
  • Each molecule in the sample material is reacted according to the reaction rule in the reaction rule set, and the reaction path corresponding to each molecule is obtained.
  • the device output product includes: Describe the sample raw materials, intermediate products and predicted products
  • calculating the first relative deviation includes:
  • the second set is not a subset of the first set, acquiring a pre-stored relative deviation value that does not meet a preset condition as the first relative deviation value;
  • the first relative deviation is calculated by the following calculation formula:
  • x 1 is the first relative deviation
  • M is the first set
  • M 1 is the set of single molecules in the molecular composition of the sample material
  • M 2 is the single molecule in the molecular composition of the intermediate product consisting of a set of species
  • M 3 of the second set, card represents the number of elements in a set.
  • reaction rate algorithm respectively calculate the reaction rate of the reaction path corresponding to each molecule in the molecular composition of the sample raw material
  • reaction rate of each reaction path is calculated according to the reaction rate constant in the reaction rate algorithm
  • k is the reaction rate constant
  • k B is the Boltzmann constant
  • h is the Planck constant
  • R is the ideal gas constant
  • E is the temperature value of the environment where the reaction path is located
  • exp is the exponent based on the natural constant Function
  • ⁇ S is the entropy change before and after the reaction corresponding to the reaction rule corresponding to the reaction path
  • ⁇ E is the reaction energy barrier corresponding to the reaction rule corresponding to the reaction path
  • P is the pressure value of the environment where the reaction path is located
  • is the pressure influence factor corresponding to the reaction rule corresponding to the reaction path.
  • the reaction rate of the reaction path is obtained according to the reaction rate constant and the reaction concentration corresponding to the reaction path.
  • the reaction rate constant has been determined, the greater the space velocity, the shorter the contact time between the raw material and the catalyst, the shorter the reaction time of the raw material, the higher the concentration of reactants in the raw material, the higher the reaction rate of the reaction path.
  • the smaller the space velocity the longer the contact time between the raw material and the catalyst, the longer the reaction time of the raw material, the lower the concentration of reactants in the raw material, and the lower the reaction rate of the reaction path.
  • the reaction rate corresponding to each reaction path is calculated by the reaction rate calculation method in the product prediction model, and combined with the single molecule content of each single molecule in the raw material, each single molecule in the predicted product can be calculated
  • the single molecule A in the raw material assumes that the single molecule A corresponds to 3 reaction paths.
  • the reaction rate corresponding to the 3 reaction paths is known.
  • the concentration of single molecule A decreases.
  • the reaction rate corresponding to the three reaction paths will decrease according to the decrease ratio of the concentration, so single molecule A will generate products in proportion to the reaction rate of the three paths.
  • the formation of each molecule can be obtained.
  • obtain the predicted product When the single molecule content of each single molecule in the catalytic reforming feed is known, the content of each single molecule in the predicted product can be obtained.
  • the calculation of the second relative deviation is, for example:
  • the second relative deviation (actual content-predicted content) ⁇ actual content.
  • the method provided by the embodiment of the present invention obtains the molecular composition of crude oil; obtains the molecular composition of different fractions obtained by distillation of the crude oil according to the physical properties of various single molecules in the molecular composition of the crude oil; according to a preset raw material ratio , Using the respective fractions as raw materials for petroleum processing, respectively input into the pre-trained product prediction model corresponding to the petroleum processing device to obtain the predicted molecular composition and the predicted molecular composition of the corresponding predicted product output by the product prediction model The predicted molecular content of each single molecule; obtaining the preset standard set of preset target products; judging whether the predicted product is based on the predicted molecular composition of the predicted product and the predicted molecular content of each single molecule in the predicted molecular composition Meet the preset standard of the target product corresponding to the predicted product in the preset standard set; if the predicted product does not meet any preset standard of the target product corresponding to the predicted product in the preset standard set Standard, the operating parameters in the product prediction model are
  • the embodiment of the present invention also provides a real-time optimization device for a molecular-level device.
  • FIG. 2 it is a structural diagram of the real-time optimization device for a molecular-level device according to an embodiment of the present invention.
  • the real-time optimization device includes: a first acquisition unit 11, a first processing unit 12, a second processing unit 13, a second acquisition unit 14, and a third processing unit 15.
  • the first processing unit 12 is used to obtain the molecular composition of different fractions obtained by distillation of the crude oil according to the physical properties of various single molecules in the molecular composition of the crude oil.
  • the second processing unit 13 is configured to use the corresponding fractions as the petroleum processing raw materials according to the preset raw material ratio, and input the pre-trained product prediction models corresponding to the petroleum processing equipment to obtain the product prediction models.
  • the predicted molecular composition of the corresponding predicted product and the predicted molecular content of each single molecule in the predicted molecular composition are output.
  • the second obtaining unit 14 is configured to obtain a preset standard set of a preset target product.
  • the third processing unit 15 judges whether the predicted product meets the target product corresponding to the predicted product in the preset standard set based on the predicted molecular composition of the predicted product and the predicted molecular content of each single molecule in the predicted molecular composition If the predicted product does not meet any of the preset standards of the target product corresponding to the predicted product in the preset standard set, adjust the operating parameters in the product prediction model to retrieve the predicted molecular composition and prediction of the predicted product The predicted molecular content of each single molecule in the molecular composition until the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set.
  • the device further includes: a flow control unit.
  • the flow control unit is used to obtain the input flow of the petroleum processing raw materials input to each petroleum processing device; determine whether each input flow meets the preset input flow range of the corresponding petroleum processing device; if any input flow does not meet the corresponding petroleum For the preset input flow range of the processing device, adjust the preset raw material ratio, and re-input the corresponding fractions as the petroleum processing raw material into the product prediction model of the corresponding petroleum processing device according to the adjusted preset raw material ratio; until each input The flow is in line with the preset input flow range of the corresponding petroleum processing device.
  • the third processing unit 15 is specifically configured to calculate the physical properties of each single molecule in the predicted molecular composition according to the predicted molecular composition of the predicted product and the predicted molecular content of each single molecule in the predicted molecular composition; The physical properties of each single molecule in the composition and the predicted molecular content of each single molecule in the predicted molecular composition are calculated to calculate the predicted physical properties of the predicted product; to determine whether the predicted physical property of each predicted product meets the prediction of the corresponding target product in the preset standard set Set physical property limit interval.
  • the device further includes: a product blending unit.
  • the product blending unit is used to blend each predicted product as a product blending raw material according to a set of preset rules to obtain the molecular composition of multiple groups of mixed products and the content of each single molecule in the mixed product; according to the content of each group of mixed products The molecular composition and the content of each single molecule in the mixed product are used to calculate the product properties of each group of mixed products.
  • the third processing unit 15 is specifically configured to determine whether the product physical properties of each group of mixed products meet the preset product physical properties of the target mixed product obtained by blending the corresponding target products in the preset standard set; if it meets If the product properties are preset, the target parameters are obtained based on all mixed products to determine whether the target parameters meet the preset conditions; if the target parameters meet the preset conditions, it is determined that the predicted product meets the prediction of the target product corresponding to the predicted product in the preset standard set.
  • the third processing unit 15 is specifically configured to obtain the product price of each group of mixed products and the output of each group of mixed products; according to the output of each group of mixed products and the product price of each group of mixed products, calculate each group Product benefits of mixed products; accumulate the product benefits of each group of mixed products to obtain cumulative benefits; obtain the raw material price of each group of petroleum processing raw materials and the operating cost of each petroleum processing device; subtract the cumulative benefits of all petroleum processing raw materials from the raw materials The price and the operating cost of all petroleum processing equipment to obtain comprehensive benefits; take the comprehensive benefits as the target parameter; judge whether the comprehensive benefit reaches the maximum value; if the comprehensive benefit reaches the maximum value, determine that the target parameter meets the preset conditions; if the comprehensive benefit does not reach the maximum value Maximum value, it is determined that the target parameter does not meet the preset conditions.
  • the third processing unit 15 is specifically configured to adjust the temperature of the environment in which the reaction path corresponding to the predicted product in the product prediction model is located; re-acquire the predicted molecular composition of the predicted product and each group of predictions according to the adjusted temperature The predicted molecular content of each single molecule in the product until the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set.
  • the third processing unit 15 is specifically configured to adjust the pressure of the environment in which the reaction path corresponding to the predicted product in the product prediction model is located; re-obtain the predicted molecular composition of the predicted product and the predicted product based on the adjusted pressure The predicted molecular content of each single molecule until the predicted product meets the preset standard of the target product corresponding to the predicted product in the preset standard set.
  • the product blending unit is specifically used to obtain the first molecular composition of each group of product blending raw materials and the first component content of each single molecule in each group of product blending raw materials; set according to preset rules , According to the first molecular composition of each group of product blending raw materials and the first component content of each single molecule in each group of product blending raw materials, the second molecular composition of each group of mixed products and each group of mixed products are obtained
  • the content of the second component of a single molecule according to the number of groups of each group contained in each single molecule in each group of mixed products and the contribution value of each group to the physical properties, calculate each single molecule in each group of mixed products
  • the physical properties of each group of mixed products calculate the physical properties of each group of mixed products based on the physical properties of each single molecule in each group of mixed products and the content of the second component.
  • the product blending unit is specifically used to obtain the number of groups of each group constituting the single molecule and the contribution value of each group to the physical properties for each single molecule; it will constitute a single molecule
  • the number of groups of each group and the contribution value of each group to the physical properties are input into the pre-trained physical property calculation model to obtain the physical properties of the single molecule output by the physical property calculation model; among them, the physical property calculation model is used to calculate the physical properties of the single molecule Calculate the physical properties of a single molecule by including the number of groups of each group and the contribution value of each group to the physical properties.
  • the device further includes:
  • the single-molecule physical property template matching unit is used to compare the group quantity of each group constituting the single molecule with the molecular information of the template single-molecule with known physical properties pre-stored in the database; the molecular information includes: the single-molecule constituting the template The number of groups for each group; determine whether there is a template single molecule that is the same as a single molecule; if there is a template single molecule that is the same as a single molecule, output the physical properties of the template single molecule as the physical properties of the single molecule; For template single molecules with the same molecule, the product blending unit performs the step of inputting the number of groups of each group constituting the single molecule and the contribution value of each group to the physical properties into a pre-trained physical property calculation model.
  • the device further includes: a model training unit.
  • the model training unit is used to construct the physical property calculation model of the single molecule; obtain the group quantity of each group constituting the sample single molecule; among them, the physical properties of the sample single molecule are known; The number of groups is input to the physical property calculation model; the predicted physical property of the sample single molecule output from the physical property calculation model is obtained; if the deviation between the predicted physical property and the known physical property is less than the preset deviation threshold, it is determined that the physical property calculation model has converged.
  • model training unit is specifically used to establish the following physical property calculation model:
  • f is the physical properties of a single molecule
  • n i is the number of groups of the i-th group in the single molecule
  • ⁇ f i is the contribution value of the i-th group in the single molecule to the physical properties
  • a is the correlation constant
  • the model training unit is specifically used to determine the number of primary groups, the number of primary groups, the number of multilevel groups, and the number of groups of multilevel groups among all groups in a single molecule; Regard all groups constituting a single molecule as primary groups; multiple groups that exist at the same time and contribute to the same physical property as multi-level groups, and the number of multiple groups as the level of multi-level groups .
  • model training unit is specifically used to establish the following physical property calculation model:
  • f is the physical properties of a single molecule
  • m 1i is the number of groups of the i-th group in the primary group
  • ⁇ f 1i is the contribution value of the i-th group in the primary group to the physical properties
  • m 2j is two The number of groups in the j-th group in the second-level group
  • ⁇ f 2j is the contribution value of the j-th group in the second-level group to the physical properties
  • m Nl is the number of the first group in the N-level group
  • ⁇ f Nl is the contribution value of the first group in the N-level group to the physical properties
  • a is the correlation constant
  • N is a positive integer greater than or equal to 2.
  • the product blending unit is specifically used in all groups of a single molecule to determine the number of primary groups, the number of primary groups, the number of multi-level groups, and the number of multi-level groups; Regard all groups constituting a single molecule as primary groups; multiple groups that exist at the same time and contribute to the same physical property as multi-level groups, and the number of multiple groups as the level of multi-level groups .
  • the product blending unit is specifically used to calculate the boiling point of a single molecule according to the following physical property calculation model:
  • T is the boiling point of a single molecule
  • SOL is a single molecule vector converted according to the number of groups of each group constituting a single molecule
  • GROUP 11 is the first contribution converted according to the contribution value of the first-level group to the boiling point.
  • Value vector GROUP 12 is the second contribution value vector converted according to the contribution value of the secondary group to the boiling point
  • GROUP 1N is the Nth contribution value vector converted according to the contribution value of the N-level group to the boiling point
  • Numh is the single The number of atoms in the molecule excluding hydrogen atoms
  • d is the first predetermined constant
  • b is the second predetermined constant
  • c is the third predetermined constant
  • N is a positive integer greater than or equal to 2.
  • the product blending unit is specifically used to calculate the density of a single molecule according to the following physical property calculation model:
  • D is the density of a single molecule
  • SOL is a single molecule vector converted according to the number of groups of each group constituting a single molecule
  • GROUP 21 is the N+th converted according to the contribution value of the first-level group to the density.
  • 1 Contribution value vector GROUP 22 is the N+2th contribution value vector transformed according to the contribution value of the secondary group to the density
  • GROUP 2N is the 2Nth contribution value vector transformed according to the N level group’s contribution to the density
  • E is the fourth preset constant
  • N is a positive integer greater than or equal to 2.
  • X is the octane number of a single molecule
  • SOL is a single molecule vector converted according to the number of groups of each group constituting a single molecule
  • GROUP 31 is converted according to the contribution value of the primary group to the octane number
  • the 2N+1 contribution value vector of the group GROUP 32 is the 2N+2 contribution value vector converted according to the contribution value of the secondary group to the octane number
  • GROUP 3N is the contribution value of the N-level group to the octane number
  • the product physical properties of the mixed product include density, cloud point, pour point, aniline point, and octane number, and of course other product physical properties. This solution will not be repeated here.
  • density is the density of the mixed product
  • D i is the density of the i-th single molecule
  • x i_volume component content is the i-th single molecule
  • the product blending unit is specifically used to calculate the cloud point contribution value of each single molecule for each group of mixed products according to the density and boiling point of each single molecule in the group of mixed products;
  • the cloud point contribution value of all single molecules in the product and the content of each single molecule are used to calculate the cloud point of the group of mixed products.
  • the product blending unit is specifically used to calculate the pour point contribution value of each single molecule according to the density and molecular weight of each single molecule in the group of mixed products for each group of mixed products;
  • the pour point contribution value of all single molecules in the product and the content of each single molecule are used to calculate the pour point of the group of mixed products.
  • the product blending unit is specifically used to calculate the aniline point contribution value of the single molecule according to the density and boiling point of the single molecule in the group of mixed products for each group of mixed products; according to all the single molecules in the group of mixed products The aniline point contribution value of the molecule and the content of each single molecule are calculated to calculate the aniline point of the mixed product.
  • the product blending unit is specifically used to obtain, for each group of mixed products, the octane number of each single molecule and the content of each single molecule in the group of mixed products; the following calculation formula is used to calculate the mixed product's octane number Octane number:
  • the ON is the octane number of the mixed product
  • HISQFG is the molecular collection
  • H is the molecular collection of normal alkanes
  • I is the molecular collection of isoalkanes
  • S is the molecular collection of cycloalkanes
  • Q is the molecular collection of olefins.
  • F is the molecular collection of aromatic hydrocarbons
  • G is the molecular collection of oxygen-containing compounds
  • ⁇ i is the content of each molecule in the mixed product
  • ⁇ H , ⁇ I , ⁇ S , ⁇ Q , ⁇ F , ⁇ G is the total content of normal paraffins, the total content of isoparaffins, the total content of cycloalkanes, the total content of olefins, the total content of aromatic hydrocarbons and the total content of oxygen-containing compounds in the mixed product respectively
  • ⁇ i regression parameters for each molecule of the mixing product ON i is the octane number of each molecule in the product mix
  • C H represents the coefficient of normal paraffins to interact with other molecules
  • C I represents isoparaffin The interaction coefficient with other molecules
  • C S represents the interaction coefficient between cycloalkanes and other molecules
  • C Q represents the interaction coefficient between olefins and other molecules
  • C F represents the interaction coefficient between aromatic hydro
  • the interaction coefficient of other molecules Represents the first constant coefficient between normal paraffin and isoparaffin, Represents the first constant coefficient between normal alkanes and cycloalkanes, Represents the first constant coefficient between normal alkanes and alkenes, Represents the first constant coefficient between normal alkanes and aromatic hydrocarbons, Represents the first constant coefficient between n-alkane and oxygen-containing compound, Represents the first constant coefficient between isoparaffin and cycloalkane, Represents the first constant coefficient between isoparaffin and olefin, Represents the first constant coefficient between isoparaffin and aromatic hydrocarbon, Represents the first constant coefficient between isoparaffin and oxygen-containing compound, Represents the first constant coefficient between cycloalkane and olefin, Represents the first constant coefficient between cycloalkane and aromatic hydrocarbon, Represents the first constant coefficient between cycloalkane and oxygen-containing compound, Represents the first constant coefficient between cycloalkane and oxygen-
  • the device further includes: a model training unit.
  • the model training unit is used to establish a product prediction model;
  • the product prediction model includes: a reaction rule set including multiple reaction rules and a reaction rate algorithm; obtain sample raw material information of the sample raw material; use the sample raw material information to set the reaction rule Perform training and fix the set of reaction rules that have been trained; use sample material information to train the reaction rate algorithm, and fix the trained reaction rate algorithm to obtain the trained product prediction model.
  • the sample raw material information of the sample raw material includes: the molecular composition of the sample raw material, the molecular content of each molecule in the sample raw material, the molecular composition of the actual product corresponding to the sample raw material, and the actual content of each molecule in the actual product .
  • the model training unit is specifically used to process the molecular composition of the sample material according to a preset reaction rule set to obtain the reaction path corresponding to each molecule in the molecular composition of the sample material; according to the molecular composition of the sample material In the reaction path corresponding to each molecule in the sample, the first molecular composition of the device output product containing the sample material, intermediate product, and predicted product is obtained; the device output product includes: sample material, intermediate product, and predicted product; according to the device output product Calculate the first relative deviation between the first molecular composition and the second molecular composition of the actual product; if the first relative deviation meets the preset conditions, the reaction rule set is fixed; if the first relative deviation does not meet the preset conditions, the reaction is adjusted According to the reaction rule in the rule set, the first relative difference is recalculated according to the adjusted reaction rule set until the first relative deviation meets the preset condition.
  • the model training unit is specifically used to obtain the types of single molecules in the first molecular composition to form the first set; to obtain the types of single molecules in the second molecular composition to form the second set; to determine whether the second set is Is a subset of the first set; if the second set is not a subset of the first set, obtain the pre-stored relative deviation value that does not meet the preset conditions as the first relative deviation value; if the second set is the first set For a subset, the first relative deviation is calculated by the following calculation formula:
  • x 1 is the first relative deviation
  • M is the first set
  • M 1 is the set of single molecules in the molecular composition of the sample material
  • M 2 is the set of single molecules in the molecular composition of the intermediate product
  • M 3 For the second set, card represents the number of elements in the set.
  • the model training unit is specifically used to calculate the reaction rate of the reaction path corresponding to each molecule in the molecular composition of the sample material according to the reaction rate algorithm; according to the molecular content and molecular weight of each molecule in the sample material According to the reaction rate corresponding to the reaction path, the predicted content of each molecule in the predicted product corresponding to the sample raw material is obtained; the second relative deviation is calculated based on the predicted content of each molecule in the predicted product and the actual content of each molecule in the actual product; if If the second relative deviation meets the preset conditions, the reaction rate algorithm is fixed; if the second relative deviation does not meet the preset conditions, adjust the parameters in the reaction rate algorithm, and recalculate the second relative deviation according to the adjusted reaction rate algorithm. Until the second relative deviation meets the preset conditions.
  • model training unit is specifically used to calculate the reaction rate of each reaction path according to the reaction rate constant in the reaction rate algorithm
  • reaction rate constant is determined according to the following calculation formula:
  • k is the reaction rate constant
  • k B is the Boltzmann constant
  • h is the Planck constant
  • R is the ideal gas constant
  • E is the temperature value of the environment where the reaction path is located
  • exp is the exponent based on the natural constant Function
  • ⁇ S is the entropy change before and after the reaction corresponding to the reaction rule corresponding to the reaction path
  • ⁇ E is the reaction energy barrier corresponding to the reaction rule corresponding to the reaction path
  • P is the pressure value of the environment where the reaction path is located
  • is the pressure influence factor corresponding to the reaction rule corresponding to the reaction path.
  • the types of petroleum processing equipment include: catalytic cracking equipment, delayed coking equipment, residual oil hydrogenation equipment, hydrocracking equipment, diesel hydro-upgrading equipment, diesel hydro-refining equipment, gasoline hydro-refining equipment , Catalytic reforming unit and alkylation unit; among them, each petroleum processing unit corresponds to a set of reaction rules.
  • the embodiment of the present invention also provides a real-time optimization system for a molecular-level device.
  • FIG. 3 it is a structural diagram of the real-time optimization system for a molecular-level device according to an embodiment of the present invention.
  • the real-time optimization system of the molecular-level device includes a processor 210 and a memory 211; the processor 210 is used to execute the real-time optimization program of the molecular-level device stored in the memory 211 to realize each
  • the method for real-time optimization of a molecular-level device described in the method embodiment includes, for example, the following steps:
  • the corresponding fractions are processed as petroleum
  • the raw materials are respectively input to the pre-trained product prediction model corresponding to the petroleum processing device to obtain the predicted molecular composition of the corresponding predicted product output by the product prediction model and the predicted molecular content of each single molecule in the predicted molecular composition;
  • the preset standard of the target product corresponding to the predicted product if the predicted product does not meet any preset standard of the target product corresponding to the predicted product in the preset standard set, adjust the product prediction model To re-acquire the predicted molecular composition of the predicted product and the predicted molecular content of each single molecule until the predicted product meets the predicted product of the target product corresponding to the predicted product in the preset standard set. Set standards.
  • the embodiment of the present invention also provides a computer-readable storage medium, the computer-readable storage medium stores one or more programs, where the storage medium may include a volatile memory, such as a random access memory; the memory is also It may include non-volatile memory, such as read-only memory, flash memory, hard disk, or solid-state memory; the memory may also include a combination of the foregoing types of memories.
  • a volatile memory such as a random access memory
  • non-volatile memory such as read-only memory, flash memory, hard disk, or solid-state memory
  • the memory may also include a combination of the foregoing types of memories.
  • the method includes the following steps:
  • the corresponding fractions are processed as petroleum
  • the raw materials are respectively input to the pre-trained product prediction model corresponding to the petroleum processing device to obtain the predicted molecular composition of the corresponding predicted product output by the product prediction model and the predicted molecular content of each single molecule in the predicted molecular composition;
  • the preset standard of the target product corresponding to the predicted product if the predicted product does not meet any preset standard of the target product corresponding to the predicted product in the preset standard set, adjust the product prediction model To re-acquire the predicted molecular composition of the predicted product and the predicted molecular content of each single molecule until the predicted product meets the predicted product of the target product corresponding to the predicted product in the preset standard set. Set standards.

Landscapes

  • Chemical & Material Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computing Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Oil, Petroleum & Natural Gas (AREA)
  • Crystallography & Structural Chemistry (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Physics & Mathematics (AREA)
  • General Chemical & Material Sciences (AREA)
  • Organic Chemistry (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Databases & Information Systems (AREA)
  • Thermal Sciences (AREA)
  • Software Systems (AREA)
  • Evolutionary Computation (AREA)
  • Data Mining & Analysis (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Medical Informatics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Production Of Liquid Hydrocarbon Mixture For Refining Petroleum (AREA)

Abstract

本发明涉及一种分子级装置的实时优化方法、装置、系统及存储介质。该方法包括:确定石油加工原料的原料分子组成;将原料分子组成输入预先训练的产物预测模型,以得到相应的预测产物的预测分子组成和每种单分子的预测分子含量;获取预先设置的目标产物的预设标准集合;根据预测分子组成和每种单分子的预测分子含量,判断预测产物是否符合预设标准集合中与预测产物对应的目标产物的预设标准;如不符合对应的目标产物的任一预设标准,则调整产物预测模型中的操作参数,以重新获取预测产物预测分子组成和每种单分子预测分子含量,直至符合预设标准。本发明实现了分子级装置从原料到产品加工过程的分子级整体模拟及实时优化,提高了精度和生产效益。

Description

一种分子级装置的实时优化方法、装置、系统及存储介质 技术领域
本发明涉及石油加工技术领域,尤其涉及一种分子级装置的实时优化方法、装置、系统及存储介质。
背景技术
在石油加工领域,现有的生产加工装置的加工过程中,要想实现生产加工装置的原料向高附加值产品高效转化,需要对原料在生产加工流程中进行最大化利用。
要实现原料在加工流程中的最大化利用,需要对生产加工装置的生成加工过程进行实时优化,然而由于生产加工装置的生成加工过程的复杂性,使得优化比较困难,难以生产出高效益的产品。
有鉴于此,如何优化生产加工装置的生产加工过程已经成为本领域技术人员急需解决的问题之一。
发明内容
为了解决上述技术问题或者至少部分地解决上述技术问题,本发明提供了一种分子级装置的实时优化方法、装置、系统及存储介质。
有鉴于此,第一方面,本发明提供了一种分子级装置的实时优化方法,所述方法包括以下步骤:
获取原油的分子组成;
根据所述原油的分子组成中各种单分子的物性,获取所述原油进行蒸馏得到的不同馏分的分子组成;
按预设原料比例,将相应的各个馏分作为石油加工原料,分别输入预先训练的与石油加工装置对应的产物预测模型,以得到所述产物预测模型输出的相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量;
获取预先设置的目标产物的预设标准集合;
根据所述预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量,判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准;
如果所述预测产物不符合所述预设标准集合中与所述预测产物对应的目标产物的任一预设标准,则调整所述产物预测模型中的操作参数,以重新获取所述预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量,直至所述预测产物符合 所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
其中,所述方法优选还包括:
获取输入每个所述石油加工装置的石油加工原料的输入流量;
判断每个所述输入流量是否均符合相应所述石油加工装置的预设输入流量范围;
若存在任一所述输入流量不符合相应所述石油加工装置的预设输入流量范围,则调整所述预设原料比例,按调整后的所述预设原料比例重新将相应的各个馏分作为石油加工原料分别输入相应的石油加工装置的产物预测模型;直至每个所述输入流量均符合相应所述石油加工装置的预设输入流量范围。
在该优选实施方式中,若每个所述输入流量均符合相应所述石油加工装置的预设输入流量范围,则认为可以进行后续步骤,执行所述得到相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量的步骤。
其中,所述判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,优选包括:
根据所述预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量计算所述预测分子组成中每种单分子的物性;
根据所述预测分子组成中每种单分子的物性和预测分子组成中每种单分子的预测分子含量,计算所述预测产物的预测物性;
判断每个所述预测产物的预测物性是否符合所述预设标准集合中对应的目标产物的预设物性限制区间。
在该优选实施方式中,若每个所述预测产物的预测物性均符合所述预设物性限制区间,则认为可以进行后续步骤,确定所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,则执行所述得到相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量的步骤。
其中,所述方法优选还包括:
将各个所述预测产物作为产品调合原料按预设规则集合进行调合,得到多组混合产品的分子组成和混合产品中每种单分子的含量;
根据每组所述混合产品的分子组成和混合产品中每种单分子的含量分别计算每组所述混合产品的产品物性。
其中,所述判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,优选包括:
判断每组所述混合产品的产品物性是否符合所述预设标准集合中对应的各个目标 产物调合得到的目标混合产品的预设产品物性;
若符合预设产品物性,则根据所有所述混合产品获取目标参数,判断所述目标参数是否符合预设条件;
若所述目标参数符合预设条件,则确定所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,并输出所述预设原料比例、产物预测模型和预设规则集合作为生产加工方案;
若所述目标参数不符合预设条件,则调整所述产物预测模型中的操作参数和所述预设规则集合中的预设规则,重新得到多组混合产品,直至每组所述混合产品的产品物性符合预设产品物性,且所有所述混合产品中的目标参数符合预设条件。
其中,所述根据所有所述混合产品获取目标参数,判断所述目标参数是否符合预设条件,优选包括:
获取每组混合产品的产品价格和每组混合产品的产量;
根据每组混合产品的产量和每组混合产品的产品价格,计算每组混合产品的产品效益;
对每组混合产品的产品效益进行累加得到累计效益;
获取每组所述石油加工原料的原料价格和每个所述石油加工装置的操作成本;
将所述累计效益减去所有石油加工原料的所述原料价格和所有石油加工装置的操作成本,得到综合效益;
将所述综合效益作为所述目标参数;
判断所述综合效益是否达到最大值;
若所述综合效益达到最大值,则确定所述目标参数符合预设条件;
若所述综合效益未达到最大值,则确定所述目标参数不符合预设条件。
其中,所述操作参数包括所述产物预测模型中反应路径所处环境的温度,所述调整所述产物预测模型中的操作参数,优选包括:
调整所述产物预测模型中与所述预测产物对应的反应路径所处环境的温度;
根据调整后的温度重新获取所述预测产物的预测分子组成和所述预测产物中每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
其中,所述操作参数包括所述产物预测模型中反应路径所处环境的压力,所述调整所述产物预测模型中的操作参数,优选包括:
调整所述产物预测模型中与所述预测产物对应的反应路径所处环境的压力;
根据调整后的压力重新获取所述预测产物的预测分子组成和所述预测产物中每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
其中,所述根据每组所述混合产品的分子组成和每种单分子的含量分别计算每组所述混合产品的产品物性,优选包括:
获取每组所述产品调合原料的第一分子组成和每组所述产品调合原料中每种单分子的第一组分含量;
按所述预设规则集合,根据每组所述产品调合原料的第一分子组成和每组所述产品调合原料中每种单分子的第一组分含量,得到每组混合产品的第二分子组成和每组所述产品调合原料中每种单分子的第二组分含量;
根据每组混合产品中每种单分子包含的每种基团的基团数量和每种基团对物性的贡献值,计算每组混合产品中每种单分子的物性;
根据每组混合产品中每种单分子的物性和第二组分含量,计算每组混合产品的物性。
其中,所述计算每种单分子的物性优选包括:
针对每种单分子,获取构成所述单分子的每种基团的基团数量,以及获取每种所述基团对物性的贡献值;
将构成所述单分子的每种基团的基团数量以及每种所述基团对物性的贡献值,输入预先训练的物性计算模型,获取所述物性计算模型输出的所述单分子的物性;其中,
所述物性计算模型,用于根据单分子包含的每种基团的基团数量以及所述每种基团对物性的贡献值,计算所述单分子的物性。
其中,所述将构成所述单分子的每种基团的基团数量以及每种所述基团对物性的贡献值,输入预先训练的物性计算模型之前,所述方法优选还包括:
将构成所述单分子的每种基团的基团数量与数据库中预存储的已知物性的模板单分子的分子信息进行比对;所述分子信息包括:构成所述模板单分子的每种基团的基团数量;
判断是否存在与所述单分子相同的所述模板单分子;
若存在与所述单分子相同的所述模板单分子,输出所述模板单分子的物性作为所述单分子的物性;
若不存在与所述单分子相同的所述模板单分子,则进行所述将构成所述单分子的每种基团的基团数量以及每种所述基团对物性的贡献值,输入预先训练的物性计算模型的 步骤。
其中,训练所述物性计算模型的步骤,优选包括:
构建单分子的物性计算模型;
获取构成样本单分子的每种基团的基团数量;其中,所述样本单分子的物性已知;
将构成所述样本单分子的每种基团的基团数量输入所述物性计算模型;
获取所述物性计算模型输出的所述样本单分子的预测物性;
如果所述预测物性与已知的所述物性之间的偏差值小于预设偏差阈值,则判定所述物性计算模型收敛,在已收敛的所述物性计算模型中获取每种基团对所述物性的贡献值,并存储所述基团对所述物性的贡献值;
如果所述预测物性与已知的所述物性之间的偏差值大于等于所述偏差阈值,则调整所述物性计算模型中每种基团对所述物性的贡献值,直到所述物性计算模型收敛为止。
其中,所述获取构成样本单分子的每种基团的基团数量,优选包括:
在所述单分子的所有基团中确定一级基团、一级基团的基团数量、多级基团和多级基团的基团数量;
将构成单分子的所有基团作为一级基团;
将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将所述多种基团的数量作为所述多级基团的级别。
其中,所述物性计算模型优选按照下述方式确定单分子的物性:
获得各种基团的基团数量与各种基团对所述物性的贡献值的乘积;
根据各种基团相应的乘积的总和获得所述单分子的物性。
例如,所述物性计算模型如下所示:f=a+∑n iΔf i
其中,f为所述单分子的物性,n i为所述单分子中第i种基团的基团数量,Δf i为所述单分子中第i种基团对所述物性的贡献值,a为关联常数。
其中,所述获取构成所述单分子的每种基团的基团数量,包括:
在所述单分子的所有基团中确定一级基团、一级基团的基团数量、多级基团和多级基团的基团数量;
将构成单分子的所有基团作为一级基团;
将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将所述多种基团的数量作为所述多级基团的级别。
其中,所述物性计算模型按照下述方式确定单分子的物性:
在每一级基团中,分别获得其所包含的各种基团的基团数量与各种基团对所述物性 的贡献值的乘积,然后获得各种基团相应的乘积的总和记为该级基团对物性的贡献值;
根据各级基团对物性的贡献值的总和获得所述单分子的物性。
例如,所述物性计算模型如下所示:
Figure PCTCN2021098570-appb-000001
其中,f为所述单分子的物性,m 1i为一级基团中第i种基团的基团数量,Δf 1i为一级基团中第i种基团对物性的贡献值,m 2j为二级基团中第j种基团的基团数量,Δf 2j为二级基团中第j种基团对物性的贡献值;m Nl为N级基团中第l种基团的基团数量,Δf Nl为N级基团中第l种基团对物性的贡献值;a为关联常数;N为大于或等于2的正整数。
其中,所述单分子的物性优选包括:单分子的沸点;
所述计算所述单分子的物性,包括:
根据如下物性计算模型计算所述单分子的沸点:
Figure PCTCN2021098570-appb-000002
其中,T为所述单分子的沸点,SOL为根据构成所述单分子的每种基团的基团数量转化得到的单分子向量,GROUP 11为根据一级基团对所述沸点的贡献值转化得到的第一贡献值向量,GROUP 12为根据二级基团对所述沸点的贡献值转化得到的第二贡献值向量,GROUP 1N为根据N级基团对沸点的贡献值转化得到的第N贡献值向量,Numh为单分子中除氢原子以外的原子个数,d为第一预设常数、b为第二预设常数、c为第三预设常数;所述N为大于或等于2的正整数。
其中,所述单分子的物性优选包括:单分子的密度;
所述物性计算模型优选按照下述方式确定所述单分子的密度:
根据构成所述单分子的每种基团的基团数量转化得到的单分子向量;
根据每一级基团对所述密度的贡献值转化得到的该级基团的贡献值向量;
获得单分子向量分别与各级基团的贡献值向量的乘积,然后获得单分子向量与各级基团相应的乘积的总和;
根据单分子向量与一级基团的贡献值向量的乘积在单分子向量与各级基团相应的乘积的总和中的占比获得所述单分子的密度。
例如,根据如下物性计算模型计算所述单分子的密度:
Figure PCTCN2021098570-appb-000003
其中,D为所述单分子的密度,SOL为根据构成所述单分子的每种基团的基团数量转化得到的单分子向量,GROUP 21为根据一级基团对所述密度的贡献值转化得到的第 N+1贡献值向量,GROUP 22为根据二级基团对所述密度的贡献值转化得到的第N+2贡献值向量,GROUP 2N为根据N级基团对密度的贡献值转化得到的第2N贡献值向量,e为第四预设常数;所述N为大于或等于2的正整数。
其中,所述单分子的物性优选包括:单分子的辛烷值;
所述物性计算模型优选按照下述方式确定单分子的辛烷值:
根据构成所述单分子的每种基团的基团数量转化得到的单分子向量;
根据每一级基团对所述辛烷值的贡献值转化得到的该级基团的贡献值向量;
获得单分子向量分别与各级基团的贡献值向量的乘积;
根据单分子向量与各级基团相应的乘积的总和获得所述单分子的辛烷值。
例如,根据如下物性计算模型计算所述单分子的辛烷值:
X=SOL×GROUP 31+SOL×GROUP 32+......+SOL×GROUP 3N+h;
其中,X为所述单分子的辛烷值,SOL为根据构成所述单分子的每种基团的基团数量转化得到的单分子向量,GROUP 31为根据一级基团对辛烷值的贡献值转化得到的第2N+1贡献值向量,GROUP 32为根据二级基团对辛烷值的贡献值转化得到的第2N+2贡献值向量,GROUP 3N为根据N级基团对辛烷值的贡献值转化得到的第3N贡献值向量;所述N为大于或等于2的正整数;h为第五预设常数。
其中,所述混合产品的产品物性,优选包括:密度、浊点、倾点、苯胺点和辛烷值。
其中,当所述混合产品的产品物性为密度时,计算每组所述混合产品的产品物性,优选包括:
通过如下方法计算每组所述混合物的密度:
针对每组混合产品,获得该组混合产品中每种单分子的密度与该种单分子的含量的乘积;
根据各种单分子相应的乘积的总和获得该组混合产品的密度。
例如,通过如下计算公式计算每组所述混合产品的密度:density=∑(D i×x i_volume);
其中,density为所述混合产品的密度,D i为第i种所述单分子的密度,x i_volume为第i种所述单分子的第二组分含量。
其中,当所述混合产品的产品物性为浊点时,计算每组所述混合产品的产品物性,优选包括:
针对每组混合产品,根据该组混合产品中每种所述单分子的密度和沸点计算得到每种所述单分子的浊点贡献值;
根据该组混合产品中所有所述单分子的浊点贡献值和每种单分子的含量,计算该组 混合产品的浊点。
其中,当所述混合产品的产品物性为倾点时,计算每组所述混合产品的产品物性,优选包括:
针对每组混合产品,根据该组混合产品中每种所述单分子的密度和分子量,计算每种所述单分子的倾点贡献值;
根据该组混合产品中所有所述单分子的倾点贡献值和每种单分子的含量,计算该组混合产品的倾点。
其中,当所述混合产品的产品物性为苯胺点时,计算每组所述混合产品的产品物性,优选包括:
针对每组混合产品,根据该组混合产品中所述单分子的密度和沸点计算得到所述单分子的苯胺点贡献值;
根据该组混合产品中所有所述单分子的苯胺点贡献值和每种单分子的含量,计算该组混合产品的苯胺点。
其中,当所述混合产品的产品物性为辛烷值时,计算每组所述混合产品的产品物性,优选包括:
针对每组混合产品,获取该组混合产品中每种所述单分子的辛烷值和每种单分子的含量;
通过如下计算公式计算所述混合产品的辛烷值:
Figure PCTCN2021098570-appb-000004
Figure PCTCN2021098570-appb-000005
Figure PCTCN2021098570-appb-000006
Figure PCTCN2021098570-appb-000007
Figure PCTCN2021098570-appb-000008
Figure PCTCN2021098570-appb-000009
Figure PCTCN2021098570-appb-000010
其中,所述ON为所述混合产品的辛烷值,HISQFG为分子集合,H为正构烷烃的分子集合,I为异构烷烃的分子集合,S为环烷烃的分子集合,Q为烯烃的分子集合,F为芳香烃的分子集合,G为含氧化合物的分子集合,υ i为所述混合产品中的各个分子的含量;υ H、υ I、υ S、υ Q、υ F、υ G分别为所述混合产品中的正构烷烃的总含量、异构烷烃的总含量、环烷烃的总含量、烯烃的总含量、芳香烃的总含量和含氧化合物的化合物总含量;β i为所述混合产品中的每种分子的回归参数;ON i为所述混合产品中的每种分子的辛烷值;C H表示正构烷烃与其他分子的交互系数;C I表示异构烷烃与其他分子的交互系数;C S表示环烷烃与其他分子的交互系数;C Q表示烯烃与其他分子的交互系数;C F表示芳香烃与其他分子的交互系数;C G表示含氧类化合物与其他分子的交互系数;
Figure PCTCN2021098570-appb-000011
表示正构烷烃与异构烷烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000012
表示正构烷烃与环烷烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000013
表示正构烷烃与烯烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000014
表示正构烷烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000015
表示正构烷烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000016
表示异构烷烃与环烷烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000017
表示异构烷烃与烯烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000018
表示异构烷烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000019
表示异构烷烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000020
表示环烷烃与烯烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000021
表示环烷烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000022
表示环烷烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000023
表示烯烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000024
表示烯烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000025
表示芳香烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000026
表示正构烷烃与异构烷烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000027
表示正构烷烃与环烷烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000028
表示正构烷烃与烯烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000029
表示正构烷烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000030
表示正构烷烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000031
表示异构烷烃与环烷烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000032
表示异构烷烃与烯烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000033
表示异构烷烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000034
表示异构烷烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000035
表示环烷烃与烯烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000036
表示环烷烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000037
表示环烷烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000038
表示烯烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000039
表示烯烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000040
表示芳香烃与含氧化合物之间的第二常数系数;其中,所述辛烷值包括:研究法辛烷值和马达法辛烷值。
其中,对产物预测模型进行训练的步骤,优选包括:
建立产物预测模型;其中,所述产物预测模型,包括:包括多种反应规则的反应规则集合以及反应速率算法;
获取样本原料的样本原料信息;
利用所述样本原料信息,对所述反应规则集合进行训练,并固定训练完成的所述反应规则集合;
利用所述样本原料信息,对所述反应速率算法进行训练,并固定训练完成的所述反应速率算法,得到训练完成的所述产物预测模型。
其中,所述样本原料的样本原料信息,优选包括:所述样本原料的分子组成,所述样本原料中每种分子的分子含量,所述样本原料对应的实际产物的分子组成以及所述实际产物中每种分子的实际含量。
其中,利用所述样本原料信息,对所述反应规则集合进行训练,优选包括:
将所述样本原料的分子组成按预设的反应规则集合进行处理,得到所述样本原料的分子组成中每种分子对应的反应路径;
根据所述样本原料的分子组成中每种分子对应的反应路径,得到包含所述样本原料、中间产物以及预测产物的装置输出产物的第一分子组成;在所述装置输出产物中,包括:所述样本原料、中间产物以及预测产物;
根据所述装置输出产物的第一分子组成与所述实际产物的第二分子组成,计算第一相对偏差;
若所述第一相对偏差符合预设条件,则固定所述反应规则集合;
若所述第一相对偏差不符合预设条件,则调整所述反应规则集合中的反应规则,根据调整后的反应规则集合,重新计算所述第一相对差值,直至所述第一相对偏差符合预设条件。
其中,根据所述装置输出产物的第一分子组成与所述实际产物的第二分子组成,计算第一相对偏差,优选包括:
获取所述第一分子组成中单分子的种类,构成第一集合;
获取所述第二分子组成中单分子的种类,构成第二集合;
判断所述第二集合是否为所述第一集合的子集;
若所述第二集合不是所述第一集合的子集,则获取预存储的不符合预设条件的相对偏差值作为所述第一相对偏差值;
若所述第二集合是所述第一集合的子集,通过如下方式计算第一相对偏差:根据所 述预测产物的分子组成中的单分子不在所述第二集合中的部分的种类数量在所述预测产物的分子组成中的单分子的种类总数量中的占比确定所述第一相对偏差。
例如,通过如下计算公式计算第一相对偏差:
Figure PCTCN2021098570-appb-000041
x 1为所述第一相对偏差,M为所述第一集合,M 1为所述样本原料的分子组成中单分子的种类组成的集合,M 2为所述中间产物的分子组成中单分子的种类组成的集合,M 3为所述第二集合,card表示集合中元素的个数。
其中,利用所述样本原料信息,对所述反应速率算法进行训练,优选包括:
根据所述反应速率算法,分别计算所述样本原料的分子组成中每种分子对应的反应路径的反应速率;
根据所述样本原料中每种分子的分子含量和所述分子的反应路径对应的反应速率,得到所述样本原料对应的预测产物中每种分子的预测含量;
根据所述预测产物中每种分子的预测含量和所述实际产物中每种分子的实际含量,计算第二相对偏差;
若所述第二相对偏差符合预设条件,则固定所述反应速率算法;
若所述第二相对偏差不符合预设条件,则调整所述反应速率算法中的参数,根据调整后的反应速率算法,重新计算所述第二相对偏差,直至所述第二相对偏差符合预设条件。
其中,根据所述反应速率算法,分别计算所述样本原料的分子组成中每种分子对应的反应路径的反应速率,优选包括:
根据所述反应速率算法中的反应速率常数,计算每条反应路径的反应速率;
其中,所述反应速率常数基于过渡态理论计算方法确定。
例如,根据如下计算公式确定所述反应速率常数:
Figure PCTCN2021098570-appb-000042
其中,k为所述反应速率常数,k B为玻尔茨曼常数,h为普朗克常数,R为理想气体常数,E为反应路径所处环境的温度值,exp为以自然常数为底的指数函数,ΔS为反应路径对应的反应规则对应的反应前后的熵变,ΔE为反应路径对应的反应规则对应的反应能垒,
Figure PCTCN2021098570-appb-000043
催化剂活性因子,P为反应路径所处环境的压力值,α为反应路径对应的反应规则对应的压力影响因子。
其中,所述获取原油进行蒸馏得到的不同馏分的分子组成,优选包括:
获取原油的分子组成;
根据所述原油的分子组成中各单分子的物性,获取所述原油进行蒸馏得到的不同馏 分的分子组成。
其中,所述石油加工装置的种类包括:
催化裂化装置,延迟焦化装置,渣油加氢装置,加氢裂化装置,柴油加氢改质装置,柴油加氢精制装置,汽油加氢精制装置,催化重整装置和烷基化装置;其中,每种石油加工装置对应一种反应规则集。
第二方面,本发明实施例提供了一种分子级装置的实时优化装置,所述实时优化装置包括:
第一获取单元,用于获取原油的分子组成;
第一处理单元,用于根据所述原油的分子组成中各种单分子的物性,获取所述原油进行蒸馏得到的不同馏分的分子组成;
第二处理单元,用于按预设原料比例,将相应的各个馏分作为石油加工原料,分别输入预先训练的与石油加工装置对应的产物预测模型,以得到所述产物预测模型输出的相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量;
第二获取单元,用于获取预先设置的目标产物的预设标准集合;
第三处理单元,根据所述预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量,判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准;如果所述预测产物不符合所述预设标准集合中与所述预测产物对应的目标产物的任一预设标准,则调整所述产物预测模型中的操作参数,以重新获取所述预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
其中,所述装置还包括:
流量控制单元,用于获取输入每个所述石油加工装置的石油加工原料的输入流量;判断每个所述输入流量是否均符合相应所述石油加工装置的预设输入流量范围;若存在任一所述输入流量不符合相应所述石油加工装置的预设输入流量范围,则调整所述预设原料比例,按调整后的所述预设原料比例重新将相应的各个馏分作为石油加工原料分别输入相应的石油加工装置的产物预测模型;直至每个所述输入流量均符合相应所述石油加工装置的预设输入流量范围。
其中,所述第三处理单元,具体用于根据所述预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量计算所述预测分子组成中每种单分子的物性;根据所述预测分子组成中每种单分子的物性和预测分子组成中每种单分子的预测分子含量,计算所述预测产物的预测物性;判断每个所述预测产物的预测物性是否符合所述预设标准 集合中对应的目标产物的预设物性限制区间。
其中,所述装置还包括:
产品调合单元,用于将各个所述预测产物作为产品调合原料按预设规则集合进行调合,得到多组混合产品的分子组成和混合产品中每种单分子的含量;根据每组所述混合产品的分子组成和混合产品中每种单分子的含量分别计算每组所述混合产品的产品物性。
其中,所述第三处理单元,具体用于判断每组所述混合产品的产品物性是否符合所述预设标准集合中对应的各个目标产物调合得到的目标混合产品的预设产品物性;
若符合预设产品物性,则根据所有所述混合产品获取目标参数,判断所述目标参数是否符合预设条件;
若所述目标参数符合预设条件,则确定所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,并输出所述预设原料比例、产物预测模型和预设规则集合作为生产加工方案;
若所述目标参数不符合预设条件,则调整所述产物预测模型中的操作参数和所述预设规则集合中的预设规则,重新得到多组混合产品,直至每组所述混合产品的产品物性符合预设产品物性,且所有所述混合产品中的目标参数符合预设条件。
其中,所述第三处理单元,具体用于获取每组混合产品的产品价格和每组混合产品的产量;根据每组混合产品的产量和每组混合产品的产品价格,计算每组混合产品的产品效益;对每组混合产品的产品效益进行累加得到累计效益;获取每组所述石油加工原料的原料价格和每个所述石油加工装置的操作成本;将所述累计效益减去所有石油加工原料的所述原料价格和所有石油加工装置的操作成本,得到综合效益;将所述综合效益作为所述目标参数;判断所述综合效益是否达到最大值;若所述综合效益达到最大值,则确定所述目标参数符合预设条件;若所述综合效益未达到最大值,则确定所述目标参数不符合预设条件。
其中,所述第三处理单元,具体用于调整所述产物预测模型中与所述预测产物对应的反应路径所处环境的温度;根据调整后的温度重新获取所述预测产物的预测分子组成和每组所述预测产物中每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
其中,所述第三处理单元,具体用于调整所述产物预测模型中与所述预测产物对应的反应路径所处环境的压力;根据调整后的压力重新获取所述预测产物的预测分子组成和所述预测产物中每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集 合中与所述预测产物对应的目标产物的预设标准。
其中,所述产品调合单元,具体用于获取每组所述产品调合原料的第一分子组成和每组所述产品调合原料中每种单分子的第一组分含量;按所述预设规则集合,根据每组所述产品调合原料的第一分子组成和每组所述产品调合原料中每种单分子的第一组分含量,得到每组混合产品的第二分子组成和每组混合产品中每种单分子的第二组分含量;根据每组混合产品中每种单分子包含的每种基团的基团数量和每种基团对物性的贡献值,计算每组混合产品中每种单分子的物性;根据每组混合产品中每种单分子的物性和所述第二组分含量,计算每组混合产品的物性。
其中,所述产品调合单元,具体用于针对每种单分子,获取构成所述单分子的每种基团的基团数量,以及获取每种所述基团对物性的贡献值;将构成所述单分子的每种基团的基团数量及每种所述基团对物性的贡献值,输入预先训练的物性计算模型,获取所述物性计算模型输出的所述单分子的物性;其中,所述物性计算模型,用于根据单分子包含的每种基团的基团数量及所述每种基团对物性的贡献值,计算所述单分子的物性。
其中,所述装置还包括:
单分子物性模板匹配单元,用于将构成所述单分子的每种基团的基团数量与数据库中预存储的已知物性的模板单分子的分子信息进行比对;所述分子信息包括:构成所述模板单分子的每种基团的基团数量;判断是否存在与所述单分子相同的所述模板单分子;若存在与所述单分子相同的所述模板单分子,输出所述模板单分子的物性作为所述单分子的物性;若不存在与所述单分子相同的所述模板单分子,则通过所述产品调合单元进行所述将构成所述单分子的每种基团的基团数量以及每种所述基团对物性的贡献值,输入预先训练的物性计算模型的步骤。
其中,所述装置还包括:
模型训练单元,用于构建单分子的物性计算模型;获取构成样本单分子的每种基团的基团数量;其中,所述样本单分子的物性已知;将构成所述样本单分子的每种基团的基团数量输入所述物性计算模型;获取所述物性计算模型输出的所述样本单分子的预测物性;如果所述预测物性与已知的所述物性之间的偏差值小于预设偏差阈值,则判定所述物性计算模型收敛,在已收敛的所述物性计算模型中获取每种基团对所述物性的贡献值,并存储所述基团对所述物性的贡献值;如果所述预测物性与已知的所述物性之间的偏差值大于等于所述偏差阈值,则调整所述物性计算模型中每种基团对所述物性的贡献值,直到所述物性计算模型收敛为止。
其中,所述模型训练单元,具体用于建立如下所示物性计算模型:f=a+∑n iΔf i
其中,f为所述单分子的物性,n i为所述单分子中第i种基团的基团数量,Δf i为所述单分子中第i种基团对所述物性的贡献值,a为关联常数。
其中,所述模型训练单元,具体用于在所述单分子的所有基团中确定一级基团、一级基团的基团数量、多级基团和多级基团的基团数量;将构成单分子的所有基团作为一级基团;将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将所述多种基团的数量作为所述多级基团的级别。
其中,所述模型训练单元,具体用于建立如下所示物性计算模型:
Figure PCTCN2021098570-appb-000044
其中,f为所述单分子的物性,m 1i为一级基团中第i种基团的基团数量,Δf 1i为一级基团中第i种基团对物性的贡献值,m 2j为二级基团中第j种基团的基团数量,Δf 2j为二级基团中第j种基团对物性的贡献值;m Nl为N级基团中第l种基团的基团数量,Δf Nl为N级基团中第l种基团对物性的贡献值;a为关联常数;N为大于或等于2的正整数。
其中,产品调合单元,具体用于所述单分子的所有基团中确定一级基团、一级基团的基团数量、多级基团和多级基团的基团数量;将构成单分子的所有基团作为一级基团;将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将所述多种基团的数量作为所述多级基团的级别。
其中,所述产品调合单元,具体用于根据如下物性计算模型计算所述单分子的沸点:
Figure PCTCN2021098570-appb-000045
其中,T为所述单分子的沸点,SOL为根据构成所述单分子的每种基团的基团数量转化得到的单分子向量,GROUP 11为根据一级基团对所述沸点的贡献值转化得到的第一贡献值向量,GROUP 12为根据二级基团对所述沸点的贡献值转化得到的第二贡献值向量,GROUP 1N为根据N级基团对沸点的贡献值转化得到的第N贡献值向量,Numh为单分子中除氢原子以外的原子个数,d为第一预设常数、b为第二预设常数、c为第三预设常数;所述N为大于或等于2的正整数。
其中,所述产品调合单元,具体用于根据如下物性计算模型计算所述单分子的密度:
Figure PCTCN2021098570-appb-000046
其中,D为所述单分子的密度,SOL为根据构成所述单分子的每种基团的基团数量转化得到的单分子向量,GROUP 21为根据一级基团对所述密度的贡献值转化得到的第N+1贡献值向量,GROUP 22为根据二级基团对所述密度的贡献值转化得到的第N+2贡 献值向量,GROUP 2N为根据N级基团对密度的贡献值转化得到的第2N贡献值向量,e为第四预设常数;所述N为大于或等于2的正整数。
其中,所述产品调合单元,具体用于根据如下物性计算模型计算所述单分子的辛烷值:X=SOL×GROUP 31+SOL×GROUP 32+......+SOL×GROUP 3N+h;
其中,X为所述单分子的辛烷值,SOL为根据构成所述单分子的每种基团的基团数量转化得到的单分子向量,GROUP 31为根据一级基团对辛烷值的贡献值转化得到的第2N+1贡献值向量,GROUP 32为根据二级基团对辛烷值的贡献值转化得到的第2N+2贡献值向量,GROUP 3N为根据N级基团对辛烷值的贡献值转化得到的第3N贡献值向量;所述N为大于或等于2的正整数;h为第五预设常数。
其中,所述混合产品的产品物性,包括:密度、浊点、倾点、苯胺点和辛烷值。
其中,所述产品调合单元,具体用于通过如下计算公式计算每组所述混合产品的密度:density=∑(D i×x i_volume);
其中,density为所述混合产品的密度,D i为第i种所述单分子的密度,x i_volume为第i种所述单分子的第二组分含量。
其中,所述产品调合单元,具体用于针对每组混合产品,根据该组混合产品中每种所述单分子的密度和沸点计算得到每种所述单分子的浊点贡献值;根据该组混合产品中所有所述单分子的浊点贡献值和每种单分子的含量,计算该组混合产品的浊点。
其中,所述产品调合单元,具体用于针对每组混合产品,根据该组混合产品中每种所述单分子的密度和分子量,计算每种所述单分子的倾点贡献值;根据该组混合产品中所有所述单分子的倾点贡献值和每种单分子的含量,计算该组混合产品的倾点。
其中,所述产品调合单元,具体用于针对每组混合产品,根据该组混合产品中所述单分子的密度和沸点计算得到所述单分子的苯胺点贡献值;根据该组混合产品中所有所述单分子的苯胺点贡献值和每种单分子的含量,计算所述混合产品的苯胺点。
其中,所述产品调合单元,具体用于针对每组混合产品,获取该组混合产品中每种所述单分子的辛烷值和每种单分子的含量;通过如下公式计算所述混合产品的辛烷值:
Figure PCTCN2021098570-appb-000047
Figure PCTCN2021098570-appb-000048
Figure PCTCN2021098570-appb-000049
Figure PCTCN2021098570-appb-000050
Figure PCTCN2021098570-appb-000051
Figure PCTCN2021098570-appb-000052
Figure PCTCN2021098570-appb-000053
其中,所述ON为所述混合产品的辛烷值,HISQFG为分子集合,H为正构烷烃的分子集合,I为异构烷烃的分子集合,S为环烷烃的分子集合,Q为烯烃的分子集合,F为芳香烃的分子集合,G为含氧化合物的分子集合,υ i为所述混合产品中的各个分子的含量;υ H、υ I、υ S、υ Q、υ F、υ G分别为所述混合产品中的正构烷烃的总含量、异构烷烃的总含量、环烷烃的总含量、烯烃的总含量、芳香烃的总含量和含氧化合物的化合物总含量;β i为所述混合产品中的每种分子的回归参数;ON i为所述混合产品中的每种分子的辛烷值;C H表示正构烷烃与其他分子的交互系数;C I表示异构烷烃与其他分子的交互系数;C S表示环烷烃与其他分子的交互系数;C Q表示烯烃与其他分子的交互系数;C F表示芳香烃与其他分子的交互系数;C G表示含氧类化合物与其他分子的交互系数;
Figure PCTCN2021098570-appb-000054
表示正构烷烃与异构烷烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000055
表示正构烷烃与环烷烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000056
表示正构烷烃与烯烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000057
表示正构烷烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000058
表示正构烷烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000059
表示异构烷烃与环烷烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000060
表示异构烷烃与烯烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000061
表示异构烷烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000062
表示异构烷烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000063
表示环烷烃与烯烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000064
表示环烷烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000065
表示环烷烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000066
表示烯烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000067
表示烯烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000068
表示芳香烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000069
表示正构烷烃与异构烷烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000070
表示正构烷烃与环烷烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000071
表示正构烷烃与烯烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000072
表示正构烷烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000073
表示正构烷烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000074
表示异构烷烃与环烷烃之间 的第二常数系数、
Figure PCTCN2021098570-appb-000075
表示异构烷烃与烯烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000076
表示异构烷烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000077
表示异构烷烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000078
表示环烷烃与烯烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000079
表示环烷烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000080
表示环烷烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000081
表示烯烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000082
表示烯烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000083
表示芳香烃与含氧化合物之间的第二常数系数;其中,所述辛烷值包括:研究法辛烷值和马达法辛烷值。
其中,所述装置还包括:
模型训练单元,用于建立产物预测模型;其中,所述产物预测模型,包括:包括多种反应规则的反应规则集合以及反应速率算法;获取样本原料的样本原料信息;利用所述样本原料信息,对所述反应规则集合进行训练,并固定训练完成的所述反应规则集合;利用所述样本原料信息,对所述反应速率算法进行训练,并固定训练完成的所述反应速率算法,得到训练完成的所述产物预测模型。
其中,所述样本原料的样本原料信息,包括:所述样本原料的分子组成,所述样本原料中每种分子的分子含量,所述样本原料对应的实际产物的分子组成以及所述实际产物中每种分子的实际含量。
其中,所述模型训练单元,具体用于将所述样本原料的分子组成按预设的反应规则集合进行处理,得到所述样本原料的分子组成中每种分子对应的反应路径;根据所述样本原料的分子组成中每种分子对应的反应路径,得到包含所述样本原料、中间产物以及预测产物的装置输出产物的第一分子组成;在所述装置输出产物中,包括:所述样本原料、中间产物以及预测产物;根据所述装置输出产物的第一分子组成与所述实际产物的第二分子组成,计算第一相对偏差;若所述第一相对偏差符合预设条件,则固定所述反应规则集合;若所述第一相对偏差不符合预设条件,则调整所述反应规则集合中的反应规则,根据调整后的反应规则集合,重新计算所述第一相对差值,直至所述第一相对偏差符合预设条件。
其中,所述模型训练单元,具体用于获取所述第一分子组成中单分子的种类,构成第一集合;获取所述第二分子组成中单分子的种类,构成第二集合;判断所述第二集合是否为所述第一集合的子集;若所述第二集合不是所述第一集合的子集,则获取预存储的不符合预设条件的相对偏差值作为所述第一相对偏差值;若所述第二集合是所述第一集合的子集,通过如下计算公式计算第一相对偏差:
Figure PCTCN2021098570-appb-000084
x 1为所述第一相对偏差,M为所述第一集合,M 1为所述样本原料的分子组成中单 分子的种类组成的集合,M 2为所述中间产物的分子组成中单分子的种类组成的集合,M 3为所述第二集合,card表示集合中元素的个数。
其中,所述模型训练单元,具体用于根据所述反应速率算法,分别计算所述样本原料的分子组成中每种分子对应的反应路径的反应速率;根据所述样本原料中每种分子的分子含量和所述分子的反应路径对应的反应速率,得到所述样本原料对应的预测产物中每种分子的预测含量;根据所述预测产物中每种分子的预测含量和所述实际产物中每种分子的实际含量,计算第二相对偏差;若所述第二相对偏差符合预设条件,则固定所述反应速率算法;若所述第二相对偏差不符合预设条件,则调整所述反应速率算法中的参数,根据调整后的反应速率算法,重新计算所述第二相对偏差,直至所述第二相对偏差符合预设条件。
其中,所述模型训练单元,具体用于根据所述反应速率算法中的反应速率常数,计算每条反应路径的反应速率;
其中,根据如下计算公式确定所述反应速率常数:
Figure PCTCN2021098570-appb-000085
其中,k为所述反应速率常数,k B为玻尔茨曼常数,h为普朗克常数,R为理想气体常数,E为反应路径所处环境的温度值,exp为以自然常数为底的指数函数,ΔS为反应路径对应的反应规则对应的反应前后的熵变,ΔE为反应路径对应的反应规则对应的反应能垒,
Figure PCTCN2021098570-appb-000086
催化剂活性因子,P为反应路径所处环境的压力值,α为反应路径对应的反应规则对应的压力影响因子。
其中,所述石油加工装置的种类包括:催化裂化装置,延迟焦化装置,渣油加氢装置,加氢裂化装置,柴油加氢改质装置,柴油加氢精制装置,汽油加氢精制装置,催化重整装置和烷基化装置;其中,每种石油加工装置对应一种反应规则集。
第三方面,本发明提供了一种分子级装置的实时优化系统,所述分子级装置的实时优化系统包括处理器、存储器;所述处理器用于执行所述存储器中存储的分子级装置的实时优化程序,以实现第一方面所述的分子级装置的实时优化方法。
第四方面,本发明提供了一种计算机可读存储介质,所述计算机可读存储介质存储有一个或者多个程序,所述一个或者多个程序可被一个或者多个处理器执行,以实现第一方面所述的分子级装置的实时优化方法。
本发明实施例提供的上述技术方案与现有技术相比具有如下优点:
本发明实施例提供的该方法,通过获取原油的分子组成;根据所述原油的分子组成中各种单分子的物性,获取所述原油进行蒸馏得到的不同馏分的分子组成;按预设原料 比例,将相应的各个馏分作为石油加工原料,分别输入预先训练的与石油加工装置对应的产物预测模型,以得到所述产物预测模型输出的相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量;获取预先设置的目标产物的预设标准集合;根据所述预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量,判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准;如果所述预测产物不符合所述预设标准集合中与所述预测产物对应的目标产物的任一预设标准,则调整所述产物预测模型中的操作参数,以重新获取所述预测产物的预测分子组成和每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。本发明实现了分子级装置从原料到产品加工过程的分子级整体模拟及实时优化,提高了精度和生产效益。
附图说明
图1为本发明实施例提供的一种分子级装置的实时优化方法的流程示意图。
图2为本发明实施例提供的一种分子级装置的实时优化装置的结构示意图。
图3为本发明另一实施例提供的一种分子级装置的实时优化系统的结构图。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚,下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例是本发明的一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有做出创造性劳动的前提下所获得的所有其他实施例,都属于本发明保护的范围。
现在将参考附图描述实现本发明各个实施例的服务器。在后续的描述中,使用用于表示元件的诸如“模块”、“部件”或“单元”的后缀仅为了有利于本发明的说明,其本身并没有特定的意义。因此,“模块”与“部件”可以混合地使用。
本发明实施例提供了一种分子级装置的实时优化方法,如图1所示,该方法可以包括以下步骤:
S101、确定石油加工原料的原料分子组成。
在本实施例中,石油加工原料的原料分子组成,即石油加工原料包括的各种分子(单分子)的信息。例如:各种单分子的种类,各种单分子的含量,各种单分子的浓度。其中,该石油加工原料的原料分子组成为基于SOL的分子组成。
在本实施例中,单分子的种类包括但不限于:烯烃、烷烃、环烷烃和芳香烃。
S102、将所述原料分子组成输入预先训练的与所述石油加工装置对应的产物预测模 型,以得到所述产物预测模型输出的相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量。
在本发明实施例中,石油加工装置的种类包括:
催化裂化装置,延迟焦化装置,渣油加氢装置,加氢裂化装置,柴油加氢改质装置,柴油加氢精制装置,汽油加氢精制装置,催化重整装置和烷基化装置;其中,每种石油加工装置对应一种反应规则集。
S103、获取预先设置的目标产物的预设标准集合。
在本发明实施例中,预设标准集合包括一个或多个预设标准,其中,预设标准包括但不限于:生成的产品的综合效益、预测产物的生成量在混合产品中的占比量、预测产物对应的预测物性。针对不同的预设标准将在后面进行描述,在此不做赘述。
S104、根据所述预测产物中的预测分子组成和每种单分子的预测分子含量,判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,如果符合,执行步骤S105;如果所述预测产物不符合所述预设标准集合中与所述预测产物对应的目标产物的任一预设标准,执行步骤S106。
S105、直接利用所述石油加工原料进行生产。
S106、调整所述产物预测模型中的操作参数,以重新获取所述预测产物的预测分子组成和每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
在本发明实施例中的操作参数包括产物预测模型中反应路径所处环境的温度、产物预测模型中反应路径所处环境的压力。针对该操作参数的调整将在后面进行描述,在此不做赘述。
在本发明另一实施例中,该方法还包括:
获取原油的分子组成;
根据所述原油的分子组成中各单分子的物性,得到所述原油进行蒸馏切割得到的不同馏分产物的分子组成;
按预设原料比例将相应的各个馏分产物作为所述石油加工原料。
在本实施例中,原油的分子种类多,不同的单分子的沸点不同,需要以不同的温度蒸馏分离,一般来说,原油中分子量越大的单分子,其沸点越高,越难分离,在原油分离过程中,按蒸馏出来的油品的种类,结合分子的沸点划分馏程,每个馏程对应一种油品,以完成对原油的分离,在本步骤中,获取原油中的单分子和每种单分子对应的含量。
在本发明实施例中,可以通过全二维气相色谱法、四级杆气相色谱-质谱仪检测法、 气相色谱/场电离-飞行时间质谱检测法、气相色谱法、近红外光谱法、核磁共振波谱法、拉曼光谱法、傅立叶变换离子回旋共振质谱法、静电场轨道阱质谱法和离子淌度质谱法中的一种或多种,确定所述石油加工原料的分子组成。当然,还可以通过其他方式来确定石油加工原料的分子组成,例如:ASTM D2425、SH/T 0606和ASTM D8144-18。
上述分子检测方法可以检测出分子的结构,并以此来得到分子的种类。但是由于原油中的分子种类数量大,虽然检测一次原油,在该原油再次使用时,可以不再检测原油的分子,但是每个单分子均进行检测的工作量大、耗时久。所以,在本方案中还可以基于结构导向集总分子表征方法构建单分子,结构导向集总分子表征方法,就是SOL分子表征方法,该方法利用24个结构增量片段表征复杂烃类分子的基础结构。任何一个石油分子都能够用一组特定的结构增量片段来表述。SOL方法属于分子尺度上的集总,将实际体系中的分子数由上百万个减少到几千,大大降低了模拟的复杂性。该表征方法不仅可以表示烷烃、环烷烃,直到包含50-60个碳原子的复杂芳烃结构,还可表示作为中间产物或二次反应产物的烯烃或环烯烃,另外还考虑了含硫、氮、氧等杂原子化合物。
在本实施例中,原油的分子组成为原油中各种分子(单分子)的信息。例如:原料包含的单分子,单分子的种类,每种单分子的体积、含量等。
在本实施例中,可以通过分别计算原油中每种单分子的沸点,基于每种单分子的沸点和含量确定馏分馏程,根据馏分馏程对原油进行蒸馏切割,得到多组馏分,在本步骤中,由于是基于单分子的物性对原油进行蒸馏,所以可以知晓原油蒸馏后得到的每组馏分的分子组成。
在本实施例中,将相应的各个馏分作为石油加工原料进行二次处理,其中预设原料比例即分别输入进不同的石油加工装置的各个馏分的比例,通过各个石油加工装置的产物预测模型,结合输入石油加工装置的馏分的分子组成,得到预测产物中的分子组成和预测产物中每种单分子的含量。
在本实施例中,原油进行蒸馏后得到的馏分包括轻质的油品和重质油品,其中轻质的油品比如石脑油,并不需要再进行二次加工,而重质油品一般需要进行不同的二次加工,使得重质油品转换为轻质的油品,以提高油品的各项性质,在本方案中,按预设原料比例将相应的各个馏分输入石油加工装置进行处理,预设原料比例包括:输入石油加工装置的馏分的种类和量,不需要进行二次加工装置的馏分则不再预设原料比例中。
在本实施例中,产物预测模型已训练优化完成,通过产物预测模型就可以得到石油加工原料在输入石油加工装置后,调整石油加工装置中的反应条件,比如压力、温度和空速等条件,以抑制某些反应的进行或者提高某些反应的进行,实现控制产物的生成, 在本步骤中,可以得到某一设定条件下的产物情况。
在本发明另一实施例中,所述方法还包括:
获取输入每个所述石油加工装置的石油加工原料的预设输入流量范围;
判断每个所述输入流量是否均符合相应所述石油加工装置的预设输入流量范围;
若存在任一所述输入流量不符合相应所述石油加工装置的预设输入流量范围,则调整所述预设原料比例,按调整后的所述预设原料比例重新将相应的各个馏分作为石油加工原料分别输入相应的石油加工装置的产物预测模型;直至每个所述输入流量均符合相应所述石油加工装置的预设输入流量范围;
若每个所述输入流量均符合相应所述石油加工装置的预设输入流量范围,则执行所述得到相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量的步骤。在本实施例中,原料的输入流量符合相应的石油加工装置的预设输入流量范围时,直接进行方案的后续步骤。
在本实施例中,每组石油加工装置均有相应的处理能力,为避免原料输入量超出石油加工装置的处理量,造成原料在石油加工装置中处理时间过短,无法完全反应的情况,更糟的情况可能会对石油加工装置造成损坏,在本实施例中,设置一预设输入流量范围,该范围的最大值可以是石油加工装置最大处理能力的百分之八十至百分之九十五之间,通过限制进入石油加工装置的原料的量,避免石油加工装置损坏。
在本实施例中,在任一石油加工装置的原料输入流量大于预设输入流量范围时,调整预设原料比例,重新规划输入石油加工装置的石油加工原料的量,使得每个石油加工装置的原料的输入流量均符合相应的石油加工装置的预设输入流量范围。
在本发明另一实施例中,上述步骤S104中的,所述判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,进一步包括:
根据所述预测分子组成中每种单分子的物性和每种单分子的预测分子含量,计算所述预测产物的预测物性;预测产物的预测物性,包括但不限于:密度、浊点、倾点、苯胺点和辛烷值。
判断每个所述预测产物的预测物性是否符合所述预设标准集合中对应的目标产物的预设物性限制区间;
若每个所述预测产物的预测物性均符合所述预设物性限制区间,则确定所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,则执行所述得到相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量的步骤。
在本发明另一实施例中,所述方法还包括:
将各个所述预测产物作为产品调合原料按预设规则集合进行调合,得到多组混合产品的分子组成和混合产品中每种单分子的含量;
根据每组所述混合产品的分子组成和每种单分子的含量分别计算每组所述混合产品的产品物性。
在本实施例中,将各个石油加工装置输入的预测产物作为产品调合原料进行调合,其中,预设规则集合中的每组预设规则中包括所使用的预测产物的种类和数量,通过混合不同石油加工装置输出的预测产物得到相应的混合产品,其中混合产品包括但不限于,用于车辆的车用油品、润滑油、液压油、齿轮油、切削油等汽油产品。通过对各个产品调合原料进行调合,使得得到的各个混合产品均符合相应产品的国家标准,即可完成生产规划。
在本实施例中,根据预测产物的分子组成和预测产物中每种单分子的含量,结合预设规则集合,得到不同的混合产品的分子组成和混合产品中每种单分子的含量。
在本发明另一实施例中,上述步骤S104中的,所述判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,进一步包括:
判断每组所述混合产品的产品物性是否符合所述预设标准集合中对应的各个目标产物调合得到的目标混合产品的预设产品物性;
若符合预设产品物性,则根据所有所述混合产品获取目标参数,判断所述目标参数是否符合预设条件;
若所述目标参数符合预设条件,则确定所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,并输出所述预设原料比例、产物预测模型和预设规则集合作为生产加工方案;
若所述目标参数不符合预设条件,则调整所述产物预测模型中的操作参数和所述预设规则集合中的预设规则,重新得到多组混合产品,直至每组所述混合产品的产品物性符合预设产品物性,且所有所述混合产品中的目标参数符合预设条件。
在本实施例中,分别计算每组混合产品的产品物性,可以通过确定每组混合产品中包含的各种单分子,即确定混合产品的分子组成,分别计算混合产品中每种单分子的物性,再根据混合汽油中每种单分子的物性和含量计算得到混合汽油产品的物性。其中,单分子的物性,包括但不限于:密度、沸点、密度、辛烷值。例如:单分子的物性还可以包括:粘度、溶解度参数、十六烷值、不饱和度等。
在本实施例中,若每组混合产品的产品物性符合预设产品物性时,说明此时调合出 来的每个混合产品均为合格产品,根据混合产品获取相关的目标参数,确认目标参数是否符合预设条件,其中,目标参数可以是产品的经济效益、产品中会对环境造成危害的物质的含量,各种混合产品中符合某项预设标准的产品在所有混合产品中的占比,在本步骤中,炼油厂进行炼油的最终目的就是为了追求效益,通过各个混合产品的价格和混合产品的量即可计算得到一个毛利值,可以通过该毛利值确认最终效益是否达到最大,来确认目标参数是否符合预设条件,确认最终效益是否达到最大可以通过随机算法进行计算,同时,随着人们环保意识的逐渐加强,混合产品中会对环境造成危害的物质的含量也会影响到该混合产品的销售,即使计算得到的效益值较大,在销售端无法售出,也无法转换成效益,所以,为增加油品的竞争力,可以对混合产品中对环境造成危害的物质的含量进行限制;同时,在不同的混合产品进行销售时,市场会有不同的需求量,比如,九十八号车用汽油的价格高于九十五号车用汽油的价格,但是九十五号车用汽油的消耗量更大,炼厂生产出大量的九十八号车用汽油,但市场的消化时间会更长,造成九十八号车用汽油库存积压,导致更多的人力和其他方面的成本,导致最终产生的效益还不如生产九十五号车用汽油,所以在本步骤中可以对符合某项预设标准的混合产品的生产量在所有混合产品中的占比,以避免产品积压。
比如,在本发明另一实施例中,上述步骤S104中的,所述判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,进一步包括:
获取每组混合产品的产品价格和每组混合产品的产量;
根据每组混合产品的产量和每组混合产品的产品价格,计算每组混合产品的产品效益;
对每组混合产品的产品效益进行累加得到累计效益;
获取每组所述石油加工原料的原料价格和每个所述石油加工装置的操作成本;
将所述累计效益减去所有石油加工原料的所述原料价格和所有石油加工装置的操作成本,得到综合效益;
将综合效益作为目标参数,判断所述综合效益是否达到最大值;
若所述综合效益达到最大值,则确定所述目标参数符合预设条件;
若所述综合效益未达到最大值,则确定所述目标参数不符合预设条件。
在本实施例中,将综合效益作为目标参数,以保证生产效益,可以通过多起点随机搜索的全局优化算法判断综合效益是否达到最大值。
在本实施例中,当目标参数也符合相应的预设条件时,说明此时整个生产流程已经符合各项生产需求,可以进行可持续性生产,此时,输出方案中的将不同的馏分输入进 不同的石油加工装置的预设原料比例、用以计算每个石油加工装置生产出的预测产物的分子组成和每种单分子的含量的产物预测模型和用以对石油加工装置产出的预测产物进行调合的预设规则集合,作为生产加工方案,在实际生产过程中,利用该生产加工方案进行生产,在分子级的层次上实现对装置的实时优化。
在本实施例中,在目标参数不符合预设条件时,说明最终调合出来的混合产品产生的经济效益可能未达到最大值,或者,混合产品中对环境造成影响的物质的量超过了设定值,或者,混合产品中符合某个预设标准的混合产品在所有混合产品中的占比量未达到设定值,此时可通过调整产物预测模型中的操作参数和预设规则集合中的预设规则,从而得到另一情况下的多组混合产品,直至本方案中输出的每组混合产品的产品物性符合预设产品物性,同时所有混合产品中的目标参数符合预设条件,即完成对分子级装置的实时优化。
在本发明另一实施例中,操作参数包括所述产物预测模型中反应路径所处环境的温度,上述步骤S106中的,调整所述产物预测模型中的操作参数,进一步包括:
调整所述产物预测模型中与所述预测产物对应的反应路径所处环境的温度;
根据调整后的温度重新获取所述预测产物的预测分子组成和每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
在本发明另一实施例中,操作参数包括所述产物预测模型中反应路径所处环境的压力,所述调整所述产物预测模型中的操作参数,进一步包括:
调整所述产物预测模型中与所述预测产物对应的反应路径所处环境的压力;
根据调整后的压力重新获取所述预测产物的预测分子组成和每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
下面对计算每组混合产品的产品物性进行进一步的描述,根据每组所述混合产品的分子组成和每种单分子的含量分别计算每组所述混合产品的产品物性,包括:
获取每组所述产品调合原料的第一分子组成和每种单分子的第一组分含量;由于产品调合原料就是每组石油加工装置的预测产物,所以,基于预测产物即可得到产品调合原料的第一分子组成和每种单分子的第一组分含量。
按所述预设规则集合,根据每组所述产品调合原料的第一分子组成和每种单分子的第一组分含量,得到每组混合产品的第二分子组成和每种单分子的第二组分含量;在本实施例中,预设规则集合中的预设规则中设定了需要的产品调合原料的种类和数量,通 过产品调合原料的分子组成和每种单分子的第一组分含量,得到混合产品的第二分子组成和每种单分子的第二组分含量。
根据每组混合产品的每种单分子包含的每种基团的基团数量和每种基团对物性的贡献值,计算每种单分子的物性;在本实施例中,针对每种单分子,获取构成单分子的每种基团的基团数量,以及获取每种基团对物性的贡献值;将构成单分子的每种基团的基团数量以及每种基团对物性的贡献值,输入预先训练的物性计算模型,获取物性计算模型输出的单分子的物性。
根据每组混合产品中每种单分子的物性和第二组分含量,计算每组混合产品的物性。
混合汽油产品的物性包括:研究法辛烷值、马达法辛烷值、雷德蒸汽压、恩氏馏程、密度、苯体积分数、芳烃体积分数、烯烃体积分数、氧质量分数和硫质量分数。
计算单分子的物性,包括:针对每种单分子,获取构成所述单分子的每种基团的基团数量,以及获取每种所述基团对物性的贡献值;
将构成所述单分子的每种基团的基团数量以及每种所述基团对物性的贡献值,输入预先训练的物性计算模型,获取所述物性计算模型输出的所述单分子的物性;其中,
所述物性计算模型,用于根据单分子包含的每种基团的基团数量以及所述每种基团对物性的贡献值,计算所述单分子的物性。
在本发明另一实施例中,所述将构成所述单分子的每种基团的基团数量以及每种所述基团对物性的贡献值,输入预先训练的物性计算模型之前,所述方法还包括:
将构成所述单分子的每种基团的基团数量与数据库中预存储的已知物性的模板单分子的分子信息进行比对;所述分子信息包括:构成所述模板单分子的每种基团的基团数量;
判断是否存在与所述单分子相同的所述模板单分子;
若存在与所述单分子相同的所述模板单分子,输出所述模板单分子的物性作为所述单分子的物性;
若不存在与所述单分子相同的所述模板单分子,则进行所述将构成所述单分子的每种基团的基团数量以及每种所述基团对物性的贡献值,输入预先训练的物性计算模型的步骤。
在本发明另一实施例中,训练所述物性计算模型的步骤,包括:
构建单分子的物性计算模型;
获取构成样本单分子的每种基团的基团数量;其中,所述样本单分子的物性已知;
将构成所述样本单分子的每种基团的基团数量输入所述物性计算模型;
获取所述物性计算模型输出的所述样本单分子的预测物性;
如果所述预测物性与已知的所述物性之间的偏差值小于预设偏差阈值,则判定所述物性计算模型收敛,在已收敛的所述物性计算模型中获取每种基团对所述物性的贡献值,并存储所述基团对所述物性的贡献值;
如果所述预测物性与已知的所述物性之间的偏差值大于等于所述偏差阈值,则调整所述物性计算模型中每种基团对所述物性的贡献值,直到所述物性计算模型收敛为止。
在本实施例中,在该物性计算模型中,包括:每种基团对物性的贡献值。该贡献值为可调的值,该贡献值在首次训练时为初始值。进一步地,在该物性计算模型中,包括:每种基团对每种物性的贡献值。
在本实施例中,预先设置训练样本集。在训练样本集中包括多个样本单分子信息。样本单分子信息,包括但不限于:构成样本单分子的每种基团的基团数量,以及样本单分子的物性。
在本实施例中,由于单分子的物性可能为多种,所以,可以在已收敛的物性计算模型中获取到每种基团分别对每种物性的贡献值。
针对每种基团而言,存储该基团对每种物性的贡献值,以便后续计算单分子的物性时,可以获取到该单分子中每种基团对需要获知的物性的贡献值,并将该单分子的每种基团的基团数量,以及每种基团对需要获知的物性的贡献值作为物性计算模型的输入,物性计算模型将该单分子的每种基团的基团数量作为模型变量,将每种基团对需要获知的物性的贡献值作为模型参量(替换物性计算模型中每种基团对物性的可调贡献值),计算出该需要获知的物性。
在本实施例中,如果样本单分子的物性为多个,那么物性计算模型输出的样本单分子的预测物性也将为多个,这时,计算每个预测物性与对应的已知物性之间的偏差值,判定所有预测物性分别与对应的已知物性之间的偏差值是否都小于预设偏差值,若是,则判定该物性计算模型收敛,根据收敛的物性计算模型中即可获取得到每种基团对应物性的贡献值,通过上述方案可以得到每种基团对不同物性的贡献值。
下面给出两种针对不同物性均可使用的物性计算模型。本领域技术人员应当知道的是,下面两种物性计算模型仅为说明本实施例,而不用于限定本实施例。
模型一、建立如下所示物性计算模型:f=a+∑n iΔf i
其中,f为所述单分子的物性,n i为所述单分子中第i种基团的基团数量,Δf i为所述单分子中第i种基团对所述物性的贡献值,a为关联常数。
例如:对于沸点而言,在基于SOL的分子表征方法中,24种基团都作为一级基团;在24种基团中,N6、N5、N4、N3、me、AA、NN、RN、NO、RO和KO等基团中的一种或者多种同时存在会对沸点存在贡献,而针对不同的物性,基团对物性的贡献值均不一致,但在不同分子中同一基团对同一物性的贡献值是一致的,基于该方案,在本实施例中构建上述物性计算模型,通过训练构建的物性计算模型,使得物性计算模型收敛,即训练模型中的每种基团对物性的贡献值,最终得到每种基团对物性的贡献值。
在本实施例中,对于构成单分子的基团,我们可以进一步划分为多级基团。进一步地,在单分子的所有基团中确定一级基团和多级基团;其中,将构成单分子的所有基团作为一级基团;将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将多种基团的数量作为多级基团的级别,我们可以根据同时存在对同一种物性会一起产生作用的多种基团作为多级基团,具体的,比如,N6和N4基团分别单独存在不同的分子中时,会对物性产生一定的影响,而其同时存在在一个分子中时,在原先对物性的贡献值的基础上,对物性的贡献值会产生一定的波动。我们划分上述多级基团的方式还可以通过基团之间的化学键力按预设键力区间进行划分,针对不同的物性,化学键力不同会有不同的影响,具体的可以根据分子稳定性对物性的影响进行划分。
在本发明另一实施例中,所述获取构成样本单分子的每种基团的基团数量,包括:
在所述单分子的所有基团中确定一级基团、一级基团的基团数量、多级基团和多级基团的基团数量;
将构成单分子的所有基团作为一级基团;
将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将所述多种基团的数量作为所述多级基团的级别。
模型二:基于划分的多级基团,可以建立如下物性计算模型:
Figure PCTCN2021098570-appb-000087
其中,f为单分子的物性,m 1i为一级基团中第i种基团的基团数量,Δf 1i为一级基团中第i种基团对物性的贡献值,m 2j为二级基团中第j种基团的基团数量,Δf 2j为二级基团中第j种基团对物性的贡献值;m Nl为N级基团中第l种基团的基团数量,Δf Nl为N级基团中第l种基团对物性的贡献值;a为关联常数;N为大于或等于2的正整数。
在本发明另一实施例中,所述获取构成所述单分子的每种基团的基团数量,包括:
在所述单分子的所有基团中确定一级基团、一级基团的基团数量、多级基团和多级基团的基团数量;
将构成单分子的所有基团作为一级基团;
将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将所述多种基团的数量作为所述多级基团的级别。
除了上述的通用型的物性计算模型,还可以根据物性种类的不同,为每种物性分别构建物性计算模型。
例如:根据如下物性计算模型计算单分子的沸点:
Figure PCTCN2021098570-appb-000088
其中,T为单分子的沸点,SOL为根据构成单分子的每种基团的基团数量转化得到的单分子向量,GROUP 1为根据一级基团对沸点的贡献值转化得到的第一贡献值向量,GROUP 2为根据二级基团对沸点的贡献值转化得到的第二贡献值向量,GROUP N为根据N级基团对沸点的贡献值转化得到的第N贡献值向量,Numh为单分子中除氢原子以外的原子个数,d为第一预设常数、b为第二预设常数、c为第三预设常数;N为大于或等于2的正整数。
根据构成单分子的每种基团的基团数量转化得到的单分子向量,包括:将构成单分子的所有基团的种类的数量作为单分子向量的维数;将每种基团的基团数量作为单分子向量中对应维度的元素值。
根据单分子的各个一级基团分别对沸点的贡献值转化得到的第一贡献值向量,包括:将一级基团的种类的数量作为第一贡献值向量的维数;将每种一级基团对沸点的贡献值作为第一贡献值向量中对应维度的元素值。根据单分子的各个二级基团分别对沸点的贡献值转化得到的第二贡献值向量,包括:将二级基团的种类的数量作为第二贡献值向量的维数;将每种二级基团对沸点的贡献值作为第二贡献值向量中对应维度的元素值。以此类推,根据单分子的各个N级基团分别对沸点的贡献值转化得到的第N贡献值向量,包括:将N级基团的种类的数量作为第N贡献值向量的维数;将每种N级基团对沸点的贡献值作为第N贡献值向量中对应维度的元素值。
又如,根据如下物性计算模型计算单分子的密度:
Figure PCTCN2021098570-appb-000089
其中,D为单分子的密度,SOL为根据构成单分子的每种基团的基团数量转化得到的单分子向量,GROUP 21为根据一级基团对密度的贡献值转化得到的第N+1贡献值向量,GROUP 22为根据二级基团对密度的贡献值转化得到的第N+2贡献值向量,GROUP 2N 为根据N级基团对密度的贡献值转化得到的第2N贡献值向量,e为第四预设常数;N为大于或等于2的正整数。
根据构成单分子的每种基团的基团数量转化得到的单分子向量,包括:将构成单分子的所有基团的种类的数量作为单分子向量的维数;将每种基团的基团数量作为单分子向量中对应维度的元素值。
根据单分子的各个一级基团分别对密度的贡献值转化得到的第N+1贡献值向量,包括:将一级基团的种类的数量作为第N+1贡献值向量的维数;将每种一级基团对密度的贡献值作为第N+1贡献值向量中对应维度的元素值。根据单分子的各个二级基团分别对密度的贡献值转化得到的第N+2贡献值向量,包括:将二级基团的种类的数量作为第N+2贡献值向量的维数;将每种二级基团对密度的贡献值作为第N+2贡献值向量中对应维度的元素值。以此类推,根据单分子的各个N级基团分别对密度的贡献值转化得到的第2N贡献值向量,包括:将N级基团的种类的数量作为第2N贡献值向量的维数;将每种N级基团对密度的贡献值作为第2N贡献值向量中对应维度的元素值。
比如,根据如下物性计算模型计算单分子的辛烷值:
X=SOL×GROUP 31+SOL×GROUP 32+......+SOL×GROUP 3N+h;
其中,X为单分子的辛烷值,SOL为根据构成单分子的每种基团的基团数量转化得到的单分子向量,GROUP 31为根据一级基团对辛烷值的贡献值转化得到的第2N+1贡献值向量,GROUP 32为根据二级基团对辛烷值的贡献值转化得到的第2N+2贡献值向量,GROUP 3N为根据N级基团对辛烷值的贡献值转化得到的第3N贡献值向量;N为大于或等于2的正整数;h为第五预设常数。
根据构成单分子的每种基团的基团数量转化得到的单分子向量,包括:将构成单分子的所有基团的种类的数量作为单分子向量的维数;将每种基团的基团数量作为单分子向量中对应维度的元素值。
根据单分子的各个一级基团分别对辛烷值的贡献值转化得到的第2N+1贡献值向量,包括:将一级基团的种类的数量作为第2N+1贡献值向量的维数;将每种一级基团对辛烷值的贡献值作为第2N+1贡献值向量中对应维度的元素值。根据单分子的各个二级基团分别对辛烷值的贡献值转化得到的第2N+2贡献值向量,包括:将二级基团的种类的数量作为第2N+2贡献值向量的维数;将每种二级基团对辛烷值的贡献值作为第2N+2贡献值向量中对应维度的元素值。以此类推,根据单分子的各个N级基团分别对辛烷值的贡献值转化得到的第3N贡献值向量,包括:将N级基团的种类的数量作为第3N贡献值向量的维数;将每种N级基团对辛烷值的贡献值作为第3N贡献值向量中对 应维度的元素值。
上述步骤中计算得到相应的单分子的物性后,将单分子作为模板单分子,并将构成单分子的每种基团的基团数量和对应的物性存储进数据库中。
混合产品的产品物性,包括:研究法辛烷值、马达法辛烷值、雷德蒸汽压、恩氏馏程、密度、苯体积分数、芳烃体积分数、烯烃体积分数、氧质量分数和硫质量分数。
下面提供五种计算混合物物性的方式,但是本领域技术人员应当知道的是,以下几种方式仅用于说明本实施例,而不用于限定本实施例。
方式一,当混合物的物性为密度时,通过如下计算公式计算混合物的密度:
density=∑(D i×x i_volume);
其中,density为混合产品的密度,D i为第i种单分子的密度,x i_volume为第i种单分子的含量。
方式二,当混合物的物性为浊点时,计算混合物的物性,包括:
根据每种单分子的密度和沸点计算得到每种单分子的浊点贡献值;
根据混合物中所有单分子的浊点贡献值和含量,计算混合物的浊点。
方式三,当混合物的物性为倾点时,计算混合物的物性,包括:
根据每种单分子的密度和分子量,计算每种单分子的倾点贡献值;
根据混合物中所有单分子的倾点贡献值和含量,计算混合物的倾点。
方式四,当混合物的物性为苯胺点时,计算混合物的物性,包括:
根据单分子的密度和沸点计算得到单分子的苯胺点贡献值;
根据混合物中所有单分子的苯胺点贡献值和含量,计算混合物的苯胺点。
方式五,当混合物的物性为辛烷值时,计算方法包括:
获取混合物中每种单分子的辛烷值和含量;
通过如下计算公式计算混合物的辛烷值:
Figure PCTCN2021098570-appb-000090
Figure PCTCN2021098570-appb-000091
Figure PCTCN2021098570-appb-000092
Figure PCTCN2021098570-appb-000093
Figure PCTCN2021098570-appb-000094
Figure PCTCN2021098570-appb-000095
Figure PCTCN2021098570-appb-000096
其中,所述ON为所述混合产品的辛烷值,HISQFG为分子集合,H为正构烷烃的分子集合,I为异构烷烃的分子集合,S为环烷烃的分子集合,Q为烯烃的分子集合,F为芳香烃的分子集合,G为含氧化合物的分子集合,υ i为所述混合产品中的各个分子的含量;υ H、υ I、υ S、υ Q、υ F、υ G分别为所述混合产品中的正构烷烃的总含量、异构烷烃的总含量、环烷烃的总含量、烯烃的总含量、芳香烃的总含量和含氧化合物的化合物总含量;β i为所述混合产品中的每种分子的回归参数;ON i为所述混合产品中的每种分子的辛烷值;C H表示正构烷烃与其他分子的交互系数;C I表示异构烷烃与其他分子的交互系数;C S表示环烷烃与其他分子的交互系数;C Q表示烯烃与其他分子的交互系数;C F表示芳香烃与其他分子的交互系数;C G表示含氧类化合物与其他分子的交互系数;
Figure PCTCN2021098570-appb-000097
表示正构烷烃与异构烷烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000098
表示正构烷烃与环烷烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000099
表示正构烷烃与烯烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000100
表示正构烷烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000101
表示正构烷烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000102
表示异构烷烃与环烷烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000103
表示异构烷烃与烯烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000104
表示异构烷烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000105
表示异构烷烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000106
表示环烷烃与烯烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000107
表示环烷烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000108
表示环烷烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000109
表示烯烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000110
表示烯烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000111
表示芳香烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000112
表示正构烷烃与异构烷烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000113
表示正构烷烃与环烷烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000114
表示正构烷烃与烯烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000115
表示正构烷烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000116
表示正构烷烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000117
表示异构烷烃与环烷烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000118
表示异构烷烃与烯烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000119
表示异构烷烃与芳 香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000120
表示异构烷烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000121
表示环烷烃与烯烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000122
表示环烷烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000123
表示环烷烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000124
表示烯烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000125
表示烯烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000126
表示芳香烃与含氧化合物之间的第二常数系数;其中,所述辛烷值包括:研究法辛烷值和马达法辛烷值。
下面对训练产物预测模型的步骤进行进一步地描述。
建立产物预测模型;其中,所述产物预测模型,包括:包括多种反应规则的反应规则集合以及反应速率算法;
获取样本原料的样本原料信息;
利用所述样本原料信息,对所述反应规则集合进行训练,并固定训练完成的所述反应规则集合;
利用所述样本原料信息,对所述反应速率算法进行训练,并固定训练完成的所述反应速率算法,得到训练完成的所述产物预测模型。
在本发明实施例中,根据石油加工装置的种类对应建立产物预测模型。
在与石油加工装置对应的产物预测模型中包括:与石油加工装置对应的反应规则集合和反应速率算法。其中,反应规则集合中包括:与石油加工装置对应的多个反应规则。
样本原料的样本原料信息,包括:样本原料的分子组成,样本原料中每种分子的分子含量,样本原料对应的实际产物的分子组成以及实际产物中每种分子的实际含量。实际产物是指样本原料经过石油加工装置处理后得到的产物。
下面给出一种对反应规则集合进行训练的方式。
将所述样本原料的分子组成按预设的反应规则集合进行处理,得到所述样本原料的分子组成中每种分子对应的反应路径;在首次计算反应路径时,将样本原料的分子组成按预设的反应规则集合进行处理,得到样本原料的分子组成中每种分子对应的反应路径。
将样本原料中的每种分子按照反应规则集合中的反应规则进行反应,得到每种分子对应的反应路径。
根据所述样本原料的分子组成中每种分子对应的反应路径,得到包含所述样本原料、中间产物以及预测产物的装置输出产物的第一分子组成;在所述装置输出产物中,包括:所述样本原料、中间产物以及预测产物;
根据所述装置输出产物的第一分子组成与所述实际产物的第二分子组成,计算第一 相对偏差;
若所述第一相对偏差符合预设条件,则固定所述反应规则集合;
若所述第一相对偏差不符合预设条件,则调整所述反应规则集合中的反应规则,根据调整后的反应规则集合,重新计算所述第一相对差值,直至所述第一相对偏差符合预设条件。
根据所述装置输出产物的第一分子组成与所述实际产物的第二分子组成,计算第一相对偏差,具体包括:
获取所述第一分子组成中单分子的种类,构成第一集合;
获取所述第二分子组成中单分子的种类,构成第二集合;
判断所述第二集合是否为所述第一集合的子集;
若所述第二集合不是所述第一集合的子集,则获取预存储的不符合预设条件的相对偏差值作为所述第一相对偏差值;
若所述第二集合是所述第一集合的子集,通过如下计算公式计算第一相对偏差:
Figure PCTCN2021098570-appb-000127
x 1为所述第一相对偏差,M为所述第一集合,M 1为所述样本原料的分子组成中单分子的种类组成的集合,M 2为所述中间产物的分子组成中单分子的种类组成的集合,M 3为所述第二集合,card表示集合中元素的个数。
下面给出一种对反应速率算法进行训练的方式。本领域技术人员应当知道的是,该方式仅用于说明本实施例,而不用于限定本实施例。
根据所述反应速率算法,分别计算所述样本原料的分子组成中每种分子对应的反应路径的反应速率;
根据所述样本原料中每种分子的分子含量和所述分子的反应路径对应的反应速率,得到所述样本原料对应的预测产物中每种分子的预测含量;
根据所述预测产物中每种分子的预测含量和所述实际产物中每种分子的实际含量,计算第二相对偏差;
若所述第二相对偏差符合预设条件,则固定所述反应速率算法;
若所述第二相对偏差不符合预设条件,则调整所述反应速率算法中的参数,根据调整后的反应速率算法,重新计算所述第二相对偏差,直至所述第二相对偏差符合预设条件。
具体的,根据反应速率算法中的反应速率常数,计算每条反应路径的反应速率;
根据如下计算公式确定反应速率常数:
Figure PCTCN2021098570-appb-000128
其中,k为反应速率常数,k B为玻尔茨曼常数,h为普朗克常数,R为理想气体常数,E为反应路径所处环境的温度值,exp为以自然常数为底的指数函数,ΔS为反应路径对应的反应规则对应的反应前后的熵变,ΔE为反应路径对应的反应规则对应的反应能垒,
Figure PCTCN2021098570-appb-000129
催化剂活性因子,P为反应路径所处环境的压力值,α为反应路径对应的反应规则对应的压力影响因子。
具体的,根据反应路径对应的反应速率常数和反应浓度,得到该反应路径的反应速率。例如:在反应速率常数已经确定的情况下,空速越大,原料和催化剂的接触时间越短,原料的反应时间越短,原料中反应物的浓度越高,该反应路径的反应速率越高;相反的,空速越小,原料和催化剂的接触时间越长,原料的反应时间越长,原料中反应物的浓度越低,该反应路径的反应速率越低。
在本实施例中,通过产物预测模型中的反应速率计算方法计算得到每条反应路径对应的反应速率,结合原料中每种单分子的单分子含量,即可计算得到预测产物中每种单分子的预测含量,比如,原料中的单分子A,假设该单分子A对应了有3条反应路径,已知3条反应路径对应的反应速率,随着反应的进行,单分子A的浓度降低,3条反应路径对应的反应速率会按浓度的下降比例而降低,所以单分子A会以3条路径的反应速率的比例生成生成物,按上述步骤,即可得到每种分子进行反应得到的生成物,并得到预测产物,在知晓催化重整原料中每种单分子的单分子含量时,即可得到预测产物中每种单分子的含量。
本实施例中,计算第二相对偏差例如是:
第二相对偏差=(实际含量—预测含量)÷实际含量。
本发明实施例提供的该方法,通过获取原油的分子组成;根据所述原油的分子组成中各种单分子的物性,获取所述原油进行蒸馏得到的不同馏分的分子组成;按预设原料比例,将相应的各个馏分作为石油加工原料,分别输入预先训练的与石油加工装置对应的产物预测模型,以得到所述产物预测模型输出的相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量;获取预先设置的目标产物的预设标准集合;根据所述预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量,判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准;如果所述预测产物不符合所述预设标准集合中与所述预测产物对应的目标产物的任一预设标准,则调整所述产物预测模型中的操作参数,以重新获取所述预测产物的 预测分子组成和每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。本发明实现了分子级装置从原料到产品加工过程的分子级整体模拟及实时优化,提高了精度和生产效益。
本发明实施例还提供了一种分子级装置的实时优化装置,如图2所示,为本发明实施例的分子级装置的实时优化装置的结构图。实时优化装置包括:第一获取单元11、第一处理单元12、第二处理单元13、第二获取单元14和第三处理单元15。
在本实施例中,第一处理单元12,用于根据原油的分子组成中各种单分子的物性,获取原油进行蒸馏得到的不同馏分的分子组成。
在本实施例中,第二处理单元13,用于按预设原料比例,将相应的各个馏分作为石油加工原料,分别输入预先训练的与石油加工装置对应的产物预测模型,以得到产物预测模型输出的相应的预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量。
在本实施例中,第二获取单元14,用于获取预先设置的目标产物的预设标准集合。
在本实施例中,第三处理单元15,根据预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量,判断预测产物是否符合预设标准集合中与预测产物对应的目标产物的预设标准;如果预测产物不符合预设标准集合中与预测产物对应的目标产物的任一预设标准,则调整产物预测模型中的操作参数,以重新获取预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量,直至预测产物符合预设标准集合中与预测产物对应的目标产物的预设标准。
在本实施例中,装置还包括:流量控制单元。
流量控制单元,用于获取输入每个石油加工装置的石油加工原料的输入流量;判断每个输入流量是否均符合相应石油加工装置的预设输入流量范围;若存在任一输入流量不符合相应石油加工装置的预设输入流量范围,则调整预设原料比例,按调整后的预设原料比例重新将相应的各个馏分作为石油加工原料分别输入相应的石油加工装置的产物预测模型;直至每个输入流量均符合相应石油加工装置的预设输入流量范围。
在本实施例中,第三处理单元15,具体用于根据预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量计算预测分子组成中每种单分子的物性;根据预测分子组成中每种单分子的物性和预测分子组成中每种单分子的预测分子含量,计算预测产物的预测物性;判断每个预测产物的预测物性是否符合预设标准集合中对应的目标产物的预设物性限制区间。
在本实施例中,装置还包括:产品调合单元。
产品调合单元,用于将各个预测产物作为产品调合原料按预设规则集合进行调合,得到多组混合产品的分子组成和混合产品中每种单分子的含量;根据每组混合产品的分子组成和混合产品中每种单分子的含量分别计算每组混合产品的产品物性。
在本实施例中,第三处理单元15,具体用于判断每组混合产品的产品物性是否符合预设标准集合中对应的各个目标产物调合得到的目标混合产品的预设产品物性;若符合预设产品物性,则根据所有混合产品获取目标参数,判断目标参数是否符合预设条件;若目标参数符合预设条件,则确定预测产物符合预设标准集合中与预测产物对应的目标产物的预设标准,并输出预设原料比例、产物预测模型和预设规则集合作为生产加工方案;若目标参数不符合预设条件,则调整产物预测模型中的操作参数和预设规则集合中的预设规则,重新得到多组混合产品,直至每组混合产品的产品物性符合预设产品物性,且所有混合产品中的目标参数符合预设条件。
在本实施例中,第三处理单元15,具体用于获取每组混合产品的产品价格和每组混合产品的产量;根据每组混合产品的产量和每组混合产品的产品价格,计算每组混合产品的产品效益;对每组混合产品的产品效益进行累加得到累计效益;获取每组石油加工原料的原料价格和每个石油加工装置的操作成本;将累计效益减去所有石油加工原料的原料价格和所有石油加工装置的操作成本,得到综合效益;将综合效益作为目标参数;判断综合效益是否达到最大值;若综合效益达到最大值,则确定目标参数符合预设条件;若综合效益未达到最大值,则确定目标参数不符合预设条件。
在本实施例中,第三处理单元15,具体用于调整产物预测模型中与预测产物对应的反应路径所处环境的温度;根据调整后的温度重新获取预测产物的预测分子组成和每组预测产物中每种单分子的预测分子含量,直至预测产物符合预设标准集合中与预测产物对应的目标产物的预设标准。
在本实施例中,第三处理单元15,具体用于调整产物预测模型中与预测产物对应的反应路径所处环境的压力;根据调整后的压力重新获取预测产物的预测分子组成和预测产物中每种单分子的预测分子含量,直至预测产物符合预设标准集合中与预测产物对应的目标产物的预设标准。
在本实施例中,产品调合单元,具体用于获取每组产品调合原料的第一分子组成和每组产品调合原料中每种单分子的第一组分含量;按预设规则集合,根据每组产品调合原料的第一分子组成和每组产品调合原料中每种单分子的第一组分含量,得到每组混合产品的第二分子组成和每组混合产品中每种单分子的第二组分含量;根据每组混合产品 中每种单分子包含的每种基团的基团数量和每种基团对物性的贡献值,计算每组混合产品中每种单分子的物性;根据每组混合产品中每种单分子的物性和第二组分含量,计算每组混合产品的物性。
在本实施例中,产品调合单元,具体用于针对每种单分子,获取构成单分子的每种基团的基团数量,以及获取每种基团对物性的贡献值;将构成单分子的每种基团的基团数量以及每种基团对物性的贡献值,输入预先训练的物性计算模型,获取物性计算模型输出的单分子的物性;其中,物性计算模型,用于根据单分子包含的每种基团的基团数量以及每种基团对物性的贡献值,计算单分子的物性。
在本实施例中,装置还包括:
单分子物性模板匹配单元,用于将构成单分子的每种基团的基团数量与数据库中预存储的已知物性的模板单分子的分子信息进行比对;分子信息包括:构成模板单分子的每种基团的基团数量;判断是否存在与单分子相同的模板单分子;若存在与单分子相同的模板单分子,输出模板单分子的物性作为单分子的物性;若不存在与单分子相同的模板单分子,则通过产品调合单元进行将构成单分子的每种基团的基团数量以及每种基团对物性的贡献值,输入预先训练的物性计算模型的步骤。
在本实施例中,装置还包括:模型训练单元。
模型训练单元,用于构建单分子的物性计算模型;获取构成样本单分子的每种基团的基团数量;其中,样本单分子的物性已知;将构成样本单分子的每种基团的基团数量输入物性计算模型;获取物性计算模型输出的样本单分子的预测物性;如果预测物性与已知的物性之间的偏差值小于预设偏差阈值,则判定物性计算模型收敛,在已收敛的物性计算模型中获取每种基团对物性的贡献值,并存储基团对物性的贡献值;如果预测物性与已知的物性之间的偏差值大于等于偏差阈值,则调整物性计算模型中每种基团对物性的贡献值,直到物性计算模型收敛为止。
在本实施例中,模型训练单元,具体用于建立如下所示物性计算模型:
f=a+∑n iΔf i
其中,f为单分子的物性,n i为单分子中第i种基团的基团数量,Δf i为单分子中第i种基团对物性的贡献值,a为关联常数。
在本实施例中,模型训练单元,具体用于在单分子的所有基团中确定一级基团、一级基团的基团数量、多级基团和多级基团的基团数量;将构成单分子的所有基团作为一级基团;将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将多种基团的数量作为多级基团的级别。
在本实施例中,模型训练单元,具体用于建立如下所示物性计算模型:
Figure PCTCN2021098570-appb-000130
其中,f为单分子的物性,m 1i为一级基团中第i种基团的基团数量,Δf 1i为一级基团中第i种基团对物性的贡献值,m 2j为二级基团中第j种基团的基团数量,Δf 2j为二级基团中第j种基团对物性的贡献值;m Nl为N级基团中第l种基团的基团数量,Δf Nl为N级基团中第l种基团对物性的贡献值;a为关联常数;N为大于或等于2的正整数。
在本实施例中,产品调合单元,具体用于单分子的所有基团中确定一级基团、一级基团的基团数量、多级基团和多级基团的基团数量;将构成单分子的所有基团作为一级基团;将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将多种基团的数量作为多级基团的级别。
在本实施例中,产品调合单元,具体用于根据如下物性计算模型计算单分子的沸点:
Figure PCTCN2021098570-appb-000131
其中,T为单分子的沸点,SOL为根据构成单分子的每种基团的基团数量转化得到的单分子向量,GROUP 11为根据一级基团对沸点的贡献值转化得到的第一贡献值向量,GROUP 12为根据二级基团对沸点的贡献值转化得到的第二贡献值向量,GROUP 1N为根据N级基团对沸点的贡献值转化得到的第N贡献值向量,Numh为单分子中除氢原子以外的原子个数,d为第一预设常数、b为第二预设常数、c为第三预设常数;N为大于或等于2的正整数。
在本实施例中,产品调合单元,具体用于根据如下物性计算模型计算单分子的密度:
Figure PCTCN2021098570-appb-000132
其中,D为单分子的密度,SOL为根据构成单分子的每种基团的基团数量转化得到的单分子向量,GROUP 21为根据一级基团对密度的贡献值转化得到的第N+1贡献值向量,GROUP 22为根据二级基团对密度的贡献值转化得到的第N+2贡献值向量,GROUP 2N为根据N级基团对密度的贡献值转化得到的第2N贡献值向量,e为第四预设常数;N为大于或等于2的正整数。
在本实施例中,产品调合单元,具体用于根据如下物性计算模型计算单分子的辛烷值:X=SOL×GROUP 31+SOL×GROUP 32+......+SOL×GROUP 3N+h;
其中,X为单分子的辛烷值,SOL为根据构成单分子的每种基团的基团数量转化得到的单分子向量,GROUP 31为根据一级基团对辛烷值的贡献值转化得到的第2N+1贡献 值向量,GROUP 32为根据二级基团对辛烷值的贡献值转化得到的第2N+2贡献值向量,GROUP 3N为根据N级基团对辛烷值的贡献值转化得到的第3N贡献值向量;N为大于或等于2的正整数;h为第五预设常数。
在本实施例中,混合产品的产品物性,包括:密度、浊点、倾点、苯胺点和辛烷值,当然还包括其他产品物性,本方案在此不再赘述。
在本实施例中,产品调合单元,具体用于通过如下计算公式计算每组混合产品的密度:density=∑(D i×x i_volume);
其中,density为混合产品的密度,D i为第i种单分子的密度,x i_volume为第i种单分子的组分含量。
在本实施例中,产品调合单元,具体用于针对每组混合产品,根据该组混合产品中每种单分子的密度和沸点计算得到每种单分子的浊点贡献值;根据该组混合产品中所有单分子的浊点贡献值和每种单分子的含量,计算该组混合产品的浊点。
在本实施例中,产品调合单元,具体用于针对每组混合产品,根据该组混合产品中每种单分子的密度和分子量,计算每种单分子的倾点贡献值;根据该组混合产品中所有单分子的倾点贡献值和每种单分子的含量,计算该组混合产品的倾点。
在本实施例中,产品调合单元,具体用于针对每组混合产品,根据该组混合产品中单分子的密度和沸点计算得到单分子的苯胺点贡献值;根据该组混合产品中所有单分子的苯胺点贡献值和每种单分子的含量,计算混合产品的苯胺点。
在本实施例中,产品调合单元,具体用于针对每组混合产品,获取该组混合产品中每种单分子的辛烷值和每种单分子的含量;通过如下计算公式计算混合产品的辛烷值:
Figure PCTCN2021098570-appb-000133
Figure PCTCN2021098570-appb-000134
Figure PCTCN2021098570-appb-000135
Figure PCTCN2021098570-appb-000136
Figure PCTCN2021098570-appb-000137
Figure PCTCN2021098570-appb-000138
Figure PCTCN2021098570-appb-000139
其中,所述ON为所述混合产品的辛烷值,HISQFG为分子集合,H为正构烷烃的分子集合,I为异构烷烃的分子集合,S为环烷烃的分子集合,Q为烯烃的分子集合,F为芳香烃的分子集合,G为含氧化合物的分子集合,υ i为所述混合产品中的各个分子的含量;υ H、υ I、υ S、υ Q、υ F、υ G分别为所述混合产品中的正构烷烃的总含量、异构烷烃的总含量、环烷烃的总含量、烯烃的总含量、芳香烃的总含量和含氧化合物的化合物总含量;β i为所述混合产品中的每种分子的回归参数;ON i为所述混合产品中的每种分子的辛烷值;C H表示正构烷烃与其他分子的交互系数;C I表示异构烷烃与其他分子的交互系数;C S表示环烷烃与其他分子的交互系数;C Q表示烯烃与其他分子的交互系数;C F表示芳香烃与其他分子的交互系数;C G表示含氧类化合物与其他分子的交互系数;
Figure PCTCN2021098570-appb-000140
表示正构烷烃与异构烷烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000141
表示正构烷烃与环烷烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000142
表示正构烷烃与烯烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000143
表示正构烷烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000144
表示正构烷烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000145
表示异构烷烃与环烷烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000146
表示异构烷烃与烯烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000147
表示异构烷烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000148
表示异构烷烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000149
表示环烷烃与烯烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000150
表示环烷烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000151
表示环烷烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000152
表示烯烃与芳香烃之间的第一常数系数、
Figure PCTCN2021098570-appb-000153
表示烯烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000154
表示芳香烃与含氧化合物之间的第一常数系数、
Figure PCTCN2021098570-appb-000155
表示正构烷烃与异构烷烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000156
表示正构烷烃与环烷烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000157
表示正构烷烃与烯烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000158
表示正构烷烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000159
表示正构烷烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000160
表示异构烷烃与环烷烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000161
表示异构烷烃与烯烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000162
表示异构烷烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000163
表示异构烷烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000164
表示环烷烃与烯烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000165
表示环烷烃与芳香烃之间的第二常数系数、
Figure PCTCN2021098570-appb-000166
表示环烷烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000167
表示烯烃与芳香烃之间的第二常 数系数、
Figure PCTCN2021098570-appb-000168
表示烯烃与含氧化合物之间的第二常数系数、
Figure PCTCN2021098570-appb-000169
表示芳香烃与含氧化合物之间的第二常数系数;其中,所述辛烷值包括:研究法辛烷值和马达法辛烷值。
在本实施例中,装置还包括:模型训练单元。
模型训练单元,用于建立产物预测模型;其中,产物预测模型,包括:包括多种反应规则的反应规则集合以及反应速率算法;获取样本原料的样本原料信息;利用样本原料信息,对反应规则集合进行训练,并固定训练完成的反应规则集合;利用样本原料信息,对反应速率算法进行训练,并固定训练完成的反应速率算法,得到训练完成的产物预测模型。
在本实施例中,样本原料的样本原料信息,包括:样本原料的分子组成,样本原料中每种分子的分子含量,样本原料对应的实际产物的分子组成以及实际产物中每种分子的实际含量。
在本实施例中,模型训练单元,具体用于将样本原料的分子组成按预设的反应规则集合进行处理,得到样本原料的分子组成中每种分子对应的反应路径;根据样本原料的分子组成中每种分子对应的反应路径,得到包含样本原料、中间产物以及预测产物的装置输出产物的第一分子组成;在装置输出产物中,包括:样本原料、中间产物以及预测产物;根据装置输出产物的第一分子组成与实际产物的第二分子组成,计算第一相对偏差;若第一相对偏差符合预设条件,则固定反应规则集合;若第一相对偏差不符合预设条件,则调整反应规则集合中的反应规则,根据调整后的反应规则集合,重新计算第一相对差值,直至第一相对偏差符合预设条件。
在本实施例中,模型训练单元,具体用于获取第一分子组成中单分子的种类,构成第一集合;获取第二分子组成中单分子的种类,构成第二集合;判断第二集合是否为第一集合的子集;若第二集合不是第一集合的子集,则获取预存储的不符合预设条件的相对偏差值作为第一相对偏差值;若第二集合是第一集合的子集,通过如下计算公式计算第一相对偏差:
Figure PCTCN2021098570-appb-000170
x 1为第一相对偏差,M为第一集合,M 1为样本原料的分子组成中单分子的种类组成的集合,M 2为中间产物的分子组成中单分子的种类组成的集合,M 3为第二集合,card表示集合中元素的个数。
在本实施例中,模型训练单元,具体用于根据反应速率算法,分别计算样本原料的分子组成中每种分子对应的反应路径的反应速率;根据样本原料中每种分子的分子含量和分子的反应路径对应的反应速率,得到样本原料对应的预测产物中每种分子的预测含 量;根据预测产物中每种分子的预测含量和实际产物中每种分子的实际含量,计算第二相对偏差;若第二相对偏差符合预设条件,则固定反应速率算法;若第二相对偏差不符合预设条件,则调整反应速率算法中的参数,根据调整后的反应速率算法,重新计算第二相对偏差,直至第二相对偏差符合预设条件。
在本实施例中,模型训练单元,具体用于根据反应速率算法中的反应速率常数,计算每条反应路径的反应速率;
其中,根据如下计算公式确定反应速率常数:
Figure PCTCN2021098570-appb-000171
其中,k为反应速率常数,k B为玻尔茨曼常数,h为普朗克常数,R为理想气体常数,E为反应路径所处环境的温度值,exp为以自然常数为底的指数函数,ΔS为反应路径对应的反应规则对应的反应前后的熵变,ΔE为反应路径对应的反应规则对应的反应能垒,
Figure PCTCN2021098570-appb-000172
催化剂活性因子,P为反应路径所处环境的压力值,α为反应路径对应的反应规则对应的压力影响因子。
在本实施例中,石油加工装置的种类包括:催化裂化装置,延迟焦化装置,渣油加氢装置,加氢裂化装置,柴油加氢改质装置,柴油加氢精制装置,汽油加氢精制装置,催化重整装置和烷基化装置;其中,每种石油加工装置对应一种反应规则集。
本发明实施例还提供了一种分子级装置的实时优化系统,如图3所示,为本发明实施例的分子级装置的实时优化系统的结构图。
在本发明实施例中,所述分子级装置的实时优化系统包括处理器210、存储器211;所述处理器210用于执行所述存储器211中存储的分子级装置的实时优化程序,以实现各方法实施例所述的分子级装置的实时优化方法,例如包括以下步骤:
通过获取原油的分子组成;根据所述原油的分子组成中各种单分子的物性,获取所述原油进行蒸馏得到的不同馏分的分子组成;按预设原料比例,将相应的各个馏分作为石油加工原料,分别输入预先训练的与石油加工装置对应的产物预测模型,以得到所述产物预测模型输出的相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量;获取预先设置的目标产物的预设标准集合;根据所述预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量,判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准;如果所述预测产物不符合所述预设标准集合中与所述预测产物对应的目标产物的任一预设标准,则调整所述产物预测模型中的操作参数,以重新获取所述预测产物的预测分子组成和每种单分子的 预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
本发明实施例还提供了一种计算机可读存储介质,所述计算机可读存储介质上存储有一个或者多个程序,其中,存储介质可以包括易失性存储器,例如随机存取存储器;存储器也可以包括非易失性存储器,例如只读存储器、快闪存储器、硬盘或固态存储器;存储器还可以包括上述种类的存储器的组合。
当存储介质中所述一个或者多个程序可被一个或者多个处理器执行时,以实现各方法实施例所述的分子级装置的实时优化方法,例如包括以下步骤:
通过获取原油的分子组成;根据所述原油的分子组成中各种单分子的物性,获取所述原油进行蒸馏得到的不同馏分的分子组成;按预设原料比例,将相应的各个馏分作为石油加工原料,分别输入预先训练的与石油加工装置对应的产物预测模型,以得到所述产物预测模型输出的相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量;获取预先设置的目标产物的预设标准集合;根据所述预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量,判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准;如果所述预测产物不符合所述预设标准集合中与所述预测产物对应的目标产物的任一预设标准,则调整所述产物预测模型中的操作参数,以重新获取所述预测产物的预测分子组成和每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
为了描述的方便,描述以上装置时以功能分为各种单元分别描述。当然,在实施本发明时可以把各单元的功能在同一个或多个软件和/或硬件中实现。
本说明书中的各个实施例均采用递进的方式描述,各个实施例之间相同相似的部分互相参见即可,每个实施例重点说明的都是与其他实施例的不同之处。尤其,对于装置或系统实施例而言,由于其基本相似于方法实施例,所以描述得比较简单,相关之处参见方法实施例的部分说明即可。以上所描述的装置及系统实施例仅仅是示意性的,其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的,作为单元显示的部件可以是或者也可以不是物理单元,即可以位于一个地方,或者也可以分布到多个网络单元上。可以根据目标的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性劳动的情况下,即可以理解并实施。
需要说明的是,在本文中,诸如“第一”和“第二”等之类的关系术语仅仅用来将 一个实体或者操作与另一个实体或操作区分开来,而不一定要求或者暗示这些实体或操作之间存在任何这种目标的关系或者顺序。而且,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
以上所述仅是本发明的具体实施方式,使本领域技术人员能够理解或实现本发明。对这些实施例的多种修改对本领域的技术人员来说将是显而易见的,本文中所定义的一般原理可以在不脱离本发明的精神或范围的情况下,在其它实施例中实现。因此,本发明将不会被限制于本文所示的这些实施例,而是要符合与本文所申请的原理和新颖特点相一致的最宽的范围。

Claims (66)

  1. 一种分子级装置的实时优化方法,其特征在于,所述方法包括:
    获取原油的分子组成;
    根据所述原油的分子组成中各种单分子的物性,获取所述原油进行蒸馏得到的不同馏分的分子组成;
    按预设原料比例,将相应的各个馏分作为石油加工原料,分别输入预先训练的与石油加工装置对应的产物预测模型,以得到所述产物预测模型输出的相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量;
    获取预先设置的目标产物的预设标准集合;
    根据所述预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量,判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准;
    如果所述预测产物不符合所述预设标准集合中与所述预测产物对应的目标产物的任一预设标准,则调整所述产物预测模型中的操作参数,以重新获取所述预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
  2. 根据权利要求1所述的方法,其特征在于,所述方法还包括:
    获取输入每个所述石油加工装置的石油加工原料的输入流量;
    判断每个所述输入流量是否均符合相应所述石油加工装置的预设输入流量范围;
    若存在任一所述输入流量不符合相应所述石油加工装置的预设输入流量范围,则调整所述预设原料比例,按调整后的所述预设原料比例重新将相应的各个馏分作为石油加工原料分别输入相应的石油加工装置的产物预测模型;直至每个所述输入流量均符合相应所述石油加工装置的预设输入流量范围。
  3. 根据权利要求1所述的方法,其特征在于,所述判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,包括:
    根据所述预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量计算所述预测分子组成中每种单分子的物性;
    根据所述预测分子组成中每种单分子的物性和预测分子组成中每种单分子的预测分子含量,计算所述预测产物的预测物性;
    判断每个所述预测产物的预测物性是否符合所述预设标准集合中对应的目标产物的预设物性限制区间。
  4. 根据权利要求1所述的方法,其特征在于,所述方法还包括:
    将各个所述预测产物作为产品调合原料按预设规则集合进行调合,得到多组混合产品的分子组成和混合产品中每种单分子的含量;
    根据每组所述混合产品的分子组成和混合产品中每种单分子的含量分别计算每组所述混合产品的产品物性。
  5. 根据权利要求4所述的方法,其特征在于,所述判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,包括:
    判断每组所述混合产品的产品物性是否符合所述预设标准集合中对应的各个目标产物调合得到的目标混合产品的预设产品物性;
    若符合预设产品物性,则根据所有所述混合产品获取目标参数,判断所述目标参数是否符合预设条件;
    若所述目标参数符合预设条件,则确定所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,并输出所述预设原料比例、产物预测模型和预设规则集合作为生产加工方案;
    若所述目标参数不符合预设条件,则调整所述产物预测模型中的操作参数和所述预设规则集合中的预设规则,重新得到多组混合产品,直至每组所述混合产品的产品物性符合预设产品物性,且所有所述混合产品中的目标参数符合预设条件。
  6. 根据权利要求5所述的方法,其特征在于,所述根据所有所述混合产品获取目标参数,判断所述目标参数是否符合预设条件,包括:
    获取每组混合产品的产品价格和每组混合产品的产量;
    根据每组混合产品的产量和每组混合产品的产品价格,计算每组混合产品的产品效益;
    对每组混合产品的产品效益进行累加得到累计效益;
    获取每组所述石油加工原料的原料价格和每个所述石油加工装置的操作成本;
    将所述累计效益减去所有石油加工原料的所述原料价格和所有石油加工装置的操作成本,得到综合效益;
    将所述综合效益作为所述目标参数;
    判断所述综合效益是否达到最大值;
    若所述综合效益达到最大值,则确定所述目标参数符合预设条件;
    若所述综合效益未达到最大值,则确定所述目标参数不符合预设条件。
  7. 根据权利要求1所述的方法,其特征在于,所述操作参数包括所述产物预测模型 中反应路径所处环境的温度;
    所述调整所述产物预测模型中的操作参数,包括:
    调整所述产物预测模型中与所述预测产物对应的反应路径所处环境的温度;
    根据调整后的温度重新获取所述预测产物的预测分子组成和每组所述预测产物中每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
  8. 根据权利要求1所述的方法,其特征在于,所述操作参数包括所述产物预测模型中反应路径所处环境的压力;
    所述调整所述产物预测模型中的操作参数,包括:
    调整所述产物预测模型中与所述预测产物对应的反应路径所处环境的压力;
    根据调整后的压力重新获取所述预测产物的预测分子组成和所述预测产物中每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
  9. 根据权利要求4所述的方法,其特征在于,所述根据每组所述混合产品的分子组成和每种单分子的含量分别计算每组所述混合产品的产品物性,包括:
    获取每组所述产品调合原料的第一分子组成和每组所述产品调合原料中每种单分子的第一组分含量;
    按所述预设规则集合,根据每组所述产品调合原料的第一分子组成和每组所述产品调合原料中每种单分子的第一组分含量,得到每组混合产品的第二分子组成和每组混合产品中每种单分子的第二组分含量;
    根据每组混合产品中每种单分子包含的每种基团的基团数量和每种基团对物性的贡献值,计算每组混合产品中每种单分子的物性;
    根据每组混合产品中每种单分子的物性和所述第二组分含量,计算每组混合产品的物性。
  10. 根据权利要求9所述的方法,其特征在于,所述单分子的物性的计算方法包括:
    针对每种单分子,获取构成所述单分子的每种基团的基团数量,以及获取每种所述基团对物性的贡献值;
    将构成所述单分子的每种基团的基团数量以及每种所述基团对物性的贡献值,输入预先训练的物性计算模型,获取所述物性计算模型输出的所述单分子的物性;其中,
    所述物性计算模型,用于根据单分子包含的每种基团的基团数量以及所述每种基团对物性的贡献值,计算所述单分子的物性。
  11. 根据权利要求10所述的方法,其特征在于,所述将构成所述单分子的每种基团的基团数量以及每种所述基团对物性的贡献值,输入预先训练的物性计算模型之前,所述方法还包括:
    将构成所述单分子的每种基团的基团数量与数据库中预存储的已知物性的模板单分子的分子信息进行比对;所述分子信息包括:构成所述模板单分子的每种基团的基团数量;
    判断是否存在与所述单分子相同的所述模板单分子;
    若存在与所述单分子相同的所述模板单分子,输出所述模板单分子的物性作为所述单分子的物性;
    若不存在与所述单分子相同的所述模板单分子,则进行所述将构成所述单分子的每种基团的基团数量以及每种所述基团对物性的贡献值,输入预先训练的物性计算模型的步骤。
  12. 根据权利要求10或11项所述的方法,其特征在于,训练所述物性计算模型的步骤,包括:
    构建单分子的物性计算模型;
    获取构成样本单分子的每种基团的基团数量;其中,所述样本单分子的物性已知;
    将构成所述样本单分子的每种基团的基团数量输入所述物性计算模型;
    获取所述物性计算模型输出的所述样本单分子的预测物性;
    如果所述预测物性与已知的所述物性之间的偏差值小于预设偏差阈值,则判定所述物性计算模型收敛,在已收敛的所述物性计算模型中获取每种基团对所述物性的贡献值,并存储所述基团对所述物性的贡献值;
    如果所述预测物性与已知的所述物性之间的偏差值大于等于所述偏差阈值,则调整所述物性计算模型中每种基团对所述物性的贡献值,直到所述物性计算模型收敛为止。
  13. 根据权利要求12所述的方法,其特征在于,建立如下所示物性计算模型:
    f=a+Σn iΔf i
    其中,f为所述单分子的物性,n i为所述单分子中第i种基团的基团数量,Δf i为所述单分子中第i种基团对所述物性的贡献值,a为关联常数。
  14. 根据权利要求12所述的方法,其特征在于,所述获取构成样本单分子的每种基团的基团数量,包括:
    在所述单分子的所有基团中确定一级基团、一级基团的基团数量、多级基团和多级基团的基团数量;
    将构成单分子的所有基团作为一级基团;
    将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将所述多种基团的数量作为所述多级基团的级别。
  15. 根据权利要求14所述的方法,其特征在于,
    建立如下所示物性计算模型:
    Figure PCTCN2021098570-appb-100001
    其中,f为所述单分子的物性,m 1i为一级基团中第i种基团的基团数量,Δf 1i为一级基团中第i种基团对物性的贡献值,m 2j为二级基团中第j种基团的基团数量,Δf 2j为二级基团中第j种基团对物性的贡献值;m Nl为N级基团中第l种基团的基团数量,Δf Nl为N级基团中第l种基团对物性的贡献值;a为关联常数;N为大于或等于2的正整数。
  16. 根据权利要求11所述的方法,其特征在于,所述获取构成所述单分子的每种基团的基团数量,包括:
    在所述单分子的所有基团中确定一级基团、一级基团的基团数量、多级基团和多级基团的基团数量;
    将构成单分子的所有基团作为一级基团;
    将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将所述多种基团的数量作为所述多级基团的级别。
  17. 根据权利要求16所述的方法,其特征在于,
    所述单分子的物性包括:单分子的沸点;
    所述计算所述单分子的物性,包括:根据如下物性计算模型计算所述单分子的沸点:
    Figure PCTCN2021098570-appb-100002
    其中,T为所述单分子的沸点,SOL为根据构成所述单分子的每种基团的基团数量转化得到的单分子向量,GROUP 11为根据一级基团对所述沸点的贡献值转化得到的第一贡献值向量,GROUP 12为根据二级基团对所述沸点的贡献值转化得到的第二贡献值向量,GROUP 1N为根据N级基团对沸点的贡献值转化得到的第N贡献值向量,Numh为单分子中除氢原子以外的原子个数,d为第一预设常数、b为第二预设常数、c为第三预设常数;所述N为大于或等于2的正整数。
  18. 根据权利要求16所述的方法,其特征在于,
    所述单分子的物性包括:单分子的密度;
    所述计算所述单分子的物性,包括:
    根据如下物性计算模型计算所述单分子的密度:
    Figure PCTCN2021098570-appb-100003
    其中,D为所述单分子的密度,SOL为根据构成所述单分子的每种基团的基团数量转化得到的单分子向量,GROUP 21为根据一级基团对所述密度的贡献值转化得到的第N+1贡献值向量,GROUP 22为根据二级基团对所述密度的贡献值转化得到的第N+2贡献值向量,GROUP 2N为根据N级基团对密度的贡献值转化得到的第2N贡献值向量,e为第四预设常数;所述N为大于或等于2的正整数。
  19. 根据权利要求16所述的方法,其特征在于,
    所述单分子的物性包括:单分子的辛烷值;
    所述计算所述单分子的物性,包括:
    根据如下物性计算模型计算所述单分子的辛烷值:
    X=SOL×GROUP 31+SOL×GROUP 32+......+SOL×GROUP 3N+h;
    其中,X为所述单分子的辛烷值,SOL为根据构成所述单分子的每种基团的基团数量转化得到的单分子向量,GROUP 31为根据一级基团对辛烷值的贡献值转化得到的第2N+1贡献值向量,GROUP 32为根据二级基团对辛烷值的贡献值转化得到的第2N+2贡献值向量,GROUP 3N为根据N级基团对辛烷值的贡献值转化得到的第3N贡献值向量;所述N为大于或等于2的正整数;h为第五预设常数。
  20. 根据权利要求4所述的方法,其特征在于,
    所述混合产品的产品物性,包括:密度、浊点、倾点、苯胺点和辛烷值。
  21. 根据权利要求20所述的方法,其特征在于,当所述混合产品的产品物性为密度时,计算每组所述混合产品的产品物性,包括:
    通过如下计算公式计算每组所述混合产品的密度:
    density=Σ(D i×x i_volume);
    其中,density为所述混合产品的密度,D i为第i种所述单分子的密度,x i_volume为第i种所述单分子的第二组分含量。
  22. 根据权利要求20所述的方法,其特征在于,当所述混合产品的产品物性为浊点时,计算每组所述混合产品的产品物性,包括:
    针对每组混合产品,根据该组混合产品中每种所述单分子的密度和沸点计算得到每种所述单分子的浊点贡献值;
    根据该组混合产品中所有所述单分子的浊点贡献值和每种单分子的含量,计算该组 混合产品的浊点。
  23. 根据权利要求20所述的方法,其特征在于,当所述混合产品的产品物性为倾点时,计算每组所述混合产品的产品物性,包括:
    针对每组混合产品,根据该组混合产品中每种所述单分子的密度和分子量,计算每种所述单分子的倾点贡献值;
    根据该组混合产品中所有所述单分子的倾点贡献值和每种单分子的含量,计算该组混合产品的倾点。
  24. 根据权利要求20所述的方法,其特征在于,当所述混合产品的产品物性为苯胺点时,计算每组所述混合产品的产品物性,包括:
    针对每组混合产品,根据该组混合产品中所述单分子的密度和沸点计算得到所述单分子的苯胺点贡献值;
    根据该组混合产品中所有所述单分子的苯胺点贡献值和每种单分子的含量,计算所述混合产品的苯胺点。
  25. 根据权利要求20所述的方法,其特征在于,当所述混合产品的产品物性为辛烷值时,计算每组所述混合产品的产品物性,包括:
    针对每组混合产品,获取该组混合产品中每种所述单分子的辛烷值和每种单分子的含量;
    通过如下计算公式计算所述混合产品的辛烷值:
    Figure PCTCN2021098570-appb-100004
    Figure PCTCN2021098570-appb-100005
    Figure PCTCN2021098570-appb-100006
    Figure PCTCN2021098570-appb-100007
    Figure PCTCN2021098570-appb-100008
    Figure PCTCN2021098570-appb-100009
    Figure PCTCN2021098570-appb-100010
    其中,所述ON为所述混合产品的辛烷值,HISQFG为分子集合,H为正构烷烃的分子集合,I为异构烷烃的分子集合,S为环烷烃的分子集合,Q为烯烃的分子集合,F为芳香烃的分子集合,G为含氧化合物的分子集合,υ i为所述混合产品中的各个分子的含量;υ H、υ I、υ S、υ Q、υ F、υ G分别为所述混合产品中的正构烷烃的总含量、异构烷烃的总含量、环烷烃的总含量、烯烃的总含量、芳香烃的总含量和含氧化合物的化合物总含量;β i为所述混合产品中的每种分子的回归参数;ON i为所述混合产品中的每种分子的辛烷值;C H表示正构烷烃与其他分子的交互系数;C I表示异构烷烃与其他分子的交互系数;C S表示环烷烃与其他分子的交互系数;C Q表示烯烃与其他分子的交互系数;C F表示芳香烃与其他分子的交互系数;C G表示含氧类化合物与其他分子的交互系数;
    Figure PCTCN2021098570-appb-100011
    表示正构烷烃与异构烷烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100012
    表示正构烷烃与环烷烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100013
    表示正构烷烃与烯烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100014
    表示正构烷烃与芳香烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100015
    表示正构烷烃与含氧化合物之间的第一常数系数、
    Figure PCTCN2021098570-appb-100016
    表示异构烷烃与环烷烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100017
    表示异构烷烃与烯烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100018
    表示异构烷烃与芳香烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100019
    表示异构烷烃与含氧化合物之间的第一常数系数、
    Figure PCTCN2021098570-appb-100020
    表示环烷烃与烯烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100021
    表示环烷烃与芳香烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100022
    表示环烷烃与含氧化合物之间的第一常数系数、
    Figure PCTCN2021098570-appb-100023
    表示烯烃与芳香烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100024
    表示烯烃与含氧化合物之间的第一常数系数、
    Figure PCTCN2021098570-appb-100025
    表示芳香烃与含氧化合物之间的第一常数系数、
    Figure PCTCN2021098570-appb-100026
    表示正构烷烃与异构烷烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100027
    表示正构烷烃与环烷烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100028
    表示正构烷烃与烯烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100029
    表示正构烷烃与芳香烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100030
    表示正构烷烃与含氧化合物之间的第二常数系数、
    Figure PCTCN2021098570-appb-100031
    表示异构烷烃与环烷烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100032
    表示异构烷烃与烯烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100033
    表示异构烷烃与芳香烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100034
    表示异构烷烃与含氧化合物之间的第二常数系数、
    Figure PCTCN2021098570-appb-100035
    表示环烷烃与烯烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100036
    表示环烷烃与芳香烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100037
    表示环烷烃与含氧化合物之间的第二常数系数、
    Figure PCTCN2021098570-appb-100038
    表示烯烃与芳香烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100039
    表示烯烃与含氧化合物之间的第二常数系数、
    Figure PCTCN2021098570-appb-100040
    表示芳香烃与含氧化合物之间的第二常数系数;其中,所述辛烷值包括:研究法辛烷值和马达法辛烷值。
  26. 根据权利要求1所述的方法,其特征在于,对产物预测模型进行训练的步骤,包括:
    建立产物预测模型;其中,所述产物预测模型,包括:包括多种反应规则的反应规则集合以及反应速率算法;
    获取样本原料的样本原料信息;
    利用所述样本原料信息,对所述反应规则集合进行训练,并固定训练完成的所述反应规则集合;
    利用所述样本原料信息,对所述反应速率算法进行训练,并固定训练完成的所述反应速率算法,得到训练完成的所述产物预测模型。
  27. 根据权利要求26所述的方法,其特征在于,所述样本原料的样本原料信息,包括:所述样本原料的分子组成,所述样本原料中每种分子的分子含量,所述样本原料对应的实际产物的分子组成以及所述实际产物中每种分子的实际含量。
  28. 根据权利要求27所述的方法,其特征在于,利用所述样本原料信息,对所述反应规则集合进行训练,包括:
    将所述样本原料的分子组成按预设的反应规则集合进行处理,得到所述样本原料的分子组成中每种分子对应的反应路径;
    根据所述样本原料的分子组成中每种分子对应的反应路径,得到包含所述样本原料、中间产物以及预测产物的装置输出产物的第一分子组成;在所述装置输出产物中,包括:所述样本原料、中间产物以及预测产物;
    根据所述装置输出产物的第一分子组成与所述实际产物的第二分子组成,计算第一相对偏差;
    若所述第一相对偏差符合预设条件,则固定所述反应规则集合;
    若所述第一相对偏差不符合预设条件,则调整所述反应规则集合中的反应规则,根据调整后的反应规则集合,重新计算所述第一相对差值,直至所述第一相对偏差符合预设条件。
  29. 根据权利要求28所述的方法,其特征在于,根据所述装置输出产物的第一分子组成与所述实际产物的第二分子组成,计算第一相对偏差,包括:
    获取所述第一分子组成中单分子的种类,构成第一集合;
    获取所述第二分子组成中单分子的种类,构成第二集合;
    判断所述第二集合是否为所述第一集合的子集;
    若所述第二集合不是所述第一集合的子集,则获取预存储的不符合预设条件的相对 偏差值作为所述第一相对偏差值;
    若所述第二集合是所述第一集合的子集,通过如下计算公式计算第一相对偏差:
    Figure PCTCN2021098570-appb-100041
    x 1为所述第一相对偏差,M为所述第一集合,M 1为所述样本原料的分子组成中单分子的种类组成的集合,M 2为所述中间产物的分子组成中单分子的种类组成的集合,M 3为所述第二集合,card表示集合中元素的个数。
  30. 根据权利要求27所述的方法,其特征在于,利用所述样本原料信息,对所述反应速率算法进行训练,包括:
    根据所述反应速率算法,分别计算所述样本原料的分子组成中每种分子对应的反应路径的反应速率;
    根据所述样本原料中每种分子的分子含量和所述分子的反应路径对应的反应速率,得到所述样本原料对应的预测产物中每种分子的预测含量;
    根据所述预测产物中每种分子的预测含量和所述实际产物中每种分子的实际含量,计算第二相对偏差;
    若所述第二相对偏差符合预设条件,则固定所述反应速率算法;
    若所述第二相对偏差不符合预设条件,则调整所述反应速率算法中的参数,根据调整后的反应速率算法,重新计算所述第二相对偏差,直至所述第二相对偏差符合预设条件。
  31. 根据权利要求30所述的方法,其特征在于,根据所述反应速率算法,分别计算所述样本原料的分子组成中每种分子对应的反应路径的反应速率,包括:
    根据所述反应速率算法中的反应速率常数,计算每条反应路径的反应速率;
    其中,根据如下计算公式确定所述反应速率常数:
    Figure PCTCN2021098570-appb-100042
    其中,k为所述反应速率常数,k B为玻尔茨曼常数,h为普朗克常数,R为理想气体常数,E为反应路径所处环境的温度值,exp为以自然常数为底的指数函数,ΔS为反应路径对应的反应规则对应的反应前后的熵变,ΔE为反应路径对应的反应规则对应的反应能垒,
    Figure PCTCN2021098570-appb-100043
    催化剂活性因子,P为反应路径所处环境的压力值,α为反应路径对应的反应规则对应的压力影响因子。
  32. 根据权利要求1-31任意一项所述的方法,其特征在于,所述石油加工装置的种类包括:
    催化裂化装置,延迟焦化装置,渣油加氢装置,加氢裂化装置,柴油加氢改质装置,柴油加氢精制装置,汽油加氢精制装置,催化重整装置和烷基化装置;其中,每种石油加工装置对应一种反应规则集。
  33. 一种分子级装置的实时优化装置,其特征在于,所述实时优化装置包括:
    第一获取单元,用于获取原油的分子组成;
    第一处理单元,用于根据所述原油的分子组成中各种单分子的物性,获取所述原油进行蒸馏得到的不同馏分的分子组成;
    第二处理单元,用于按预设原料比例,将相应的各个馏分作为石油加工原料,分别输入预先训练的与石油加工装置对应的产物预测模型,以得到所述产物预测模型输出的相应的预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量;
    第二获取单元,用于获取预先设置的目标产物的预设标准集合;
    第三处理单元,根据所述预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量,判断所述预测产物是否符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准;如果所述预测产物不符合所述预设标准集合中与所述预测产物对应的目标产物的任一预设标准,则调整所述产物预测模型中的操作参数,以重新获取所述预测产物的预测分子组成和所述预测分子组成中每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
  34. 根据权利要求33所述的装置,其特征在于,所述装置还包括:
    流量控制单元,用于获取输入每个所述石油加工装置的石油加工原料的输入流量;判断每个所述输入流量是否均符合相应所述石油加工装置的预设输入流量范围;若存在任一所述输入流量不符合相应所述石油加工装置的预设输入流量范围,则调整所述预设原料比例,按调整后的所述预设原料比例重新将相应的各个馏分作为石油加工原料分别输入相应的石油加工装置的产物预测模型;直至每个所述输入流量均符合相应所述石油加工装置的预设输入流量范围。
  35. 根据权利要求33所述的装置,其特征在于,所述第三处理单元,具体用于根据所述预测产物的预测分子组成和预测分子组成中每种单分子的预测分子含量计算所述预测分子组成中每种单分子的物性;根据所述预测分子组成中每种单分子的物性和预测分子组成中每种单分子的预测分子含量,计算所述预测产物的预测物性;判断每个所述预测产物的预测物性是否符合所述预设标准集合中对应的目标产物的预设物性限制区间。
  36. 根据权利要求33所述的装置,其特征在于,所述装置还包括:
    产品调合单元,用于将各个所述预测产物作为产品调合原料按预设规则集合进行调合,得到多组混合产品的分子组成和混合产品中每种单分子的含量;根据每组所述混合产品的分子组成和混合产品中每种单分子的含量分别计算每组所述混合产品的产品物性。
  37. 根据权利要求36所述的装置,其特征在于,所述第三处理单元,具体用于判断每组所述混合产品的产品物性是否符合所述预设标准集合中对应的各个目标产物调合得到的目标混合产品的预设产品物性;
    若符合预设产品物性,则根据所有所述混合产品获取目标参数,判断所述目标参数是否符合预设条件;
    若所述目标参数符合预设条件,则确定所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准,并输出所述预设原料比例、产物预测模型和预设规则集合作为生产加工方案;
    若所述目标参数不符合预设条件,则调整所述产物预测模型中的操作参数和所述预设规则集合中的预设规则,重新得到多组混合产品,直至每组所述混合产品的产品物性符合预设产品物性,且所有所述混合产品中的目标参数符合预设条件。
  38. 根据权利要求37所述的装置,其特征在于,所述第三处理单元,具体用于获取每组混合产品的产品价格和每组混合产品的产量;根据每组混合产品的产量和每组混合产品的产品价格,计算每组混合产品的产品效益;对每组混合产品的产品效益进行累加得到累计效益;获取每组所述石油加工原料的原料价格和每个所述石油加工装置的操作成本;将所述累计效益减去所有石油加工原料的所述原料价格和所有石油加工装置的操作成本,得到综合效益;将所述综合效益作为所述目标参数;判断所述综合效益是否达到最大值;若所述综合效益达到最大值,则确定所述目标参数符合预设条件;若所述综合效益未达到最大值,则确定所述目标参数不符合预设条件。
  39. 根据权利要求33所述的装置,其特征在于,所述第三处理单元,具体用于调整所述产物预测模型中与所述预测产物对应的反应路径所处环境的温度;根据调整后的温度重新获取所述预测产物的预测分子组成和每组所述预测产物中每种单分子的预测分子含量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
  40. 根据权利要求33所述的装置,其特征在于,所述第三处理单元,具体用于调整所述产物预测模型中与所述预测产物对应的反应路径所处环境的压力;根据调整后的压力重新获取所述预测产物的预测分子组成和所述预测产物中每种单分子的预测分子含 量,直至所述预测产物符合所述预设标准集合中与所述预测产物对应的目标产物的预设标准。
  41. 根据权利要求36所述的装置,其特征在于,所述产品调合单元,具体用于获取每组所述产品调合原料的第一分子组成和每组所述产品调合原料中每种单分子的第一组分含量;按所述预设规则集合,根据每组所述产品调合原料的第一分子组成和每组所述产品调合原料中每种单分子的第一组分含量,得到每组混合产品的第二分子组成和每组混合产品中每种单分子的第二组分含量;根据每组混合产品中每种单分子包含的每种基团的基团数量和每种基团对物性的贡献值,计算每组混合产品中每种单分子的物性;根据每组混合产品中每种单分子的物性和所述第二组分含量,计算每组混合产品的物性。
  42. 根据权利要求41所述的装置,其特征在于,所述产品调合单元,具体用于针对每种单分子,获取构成所述单分子的每种基团的基团数量,以及获取每种所述基团对物性的贡献值;将构成所述单分子的每种基团的基团数量以及每种所述基团对物性的贡献值,输入预先训练的物性计算模型,获取所述物性计算模型输出的所述单分子的物性;其中,所述物性计算模型,用于根据单分子包含的每种基团的基团数量以及所述每种基团对物性的贡献值,计算所述单分子的物性。
  43. 根据权利要求42所述的装置,其特征在于,所述装置还包括:
    单分子物性模板匹配单元,用于将构成所述单分子的每种基团的基团数量与数据库中预存储的已知物性的模板单分子的分子信息进行比对;所述分子信息包括:构成所述模板单分子的每种基团的基团数量;判断是否存在与所述单分子相同的所述模板单分子;若存在与所述单分子相同的所述模板单分子,输出所述模板单分子的物性作为所述单分子的物性;若不存在与所述单分子相同的所述模板单分子,则通过所述产品调合单元进行所述将构成所述单分子的每种基团的基团数量以及每种所述基团对物性的贡献值,输入预先训练的物性计算模型的步骤。
  44. 根据权利要求42或43所述的装置,其特征在于,所述装置还包括:
    模型训练单元,用于构建单分子的物性计算模型;获取构成样本单分子的每种基团的基团数量;其中,所述样本单分子的物性已知;将构成所述样本单分子的每种基团的基团数量输入所述物性计算模型;获取所述物性计算模型输出的所述样本单分子的预测物性;如果所述预测物性与已知的所述物性之间的偏差值小于预设偏差阈值,则判定所述物性计算模型收敛,在已收敛的所述物性计算模型中获取每种基团对所述物性的贡献值,并存储所述基团对所述物性的贡献值;如果所述预测物性与已知的所述物性之间的 偏差值大于等于所述偏差阈值,则调整所述物性计算模型中每种基团对所述物性的贡献值,直到所述物性计算模型收敛为止。
  45. 根据权利要求44所述的装置,其特征在于,所述模型训练单元,具体用于建立如下所示物性计算模型:
    f=a+Σn iΔf i
    其中,f为所述单分子的物性,n i为所述单分子中第i种基团的基团数量,Δf i为所述单分子中第i种基团对所述物性的贡献值,a为关联常数。
  46. 根据权利要求44所述的装置,其特征在于,所述模型训练单元,具体用于在所述单分子的所有基团中确定一级基团、一级基团的基团数量、多级基团和多级基团的基团数量;将构成单分子的所有基团作为一级基团;将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将所述多种基团的数量作为所述多级基团的级别。
  47. 根据权利要求46所述的装置,其特征在于,所述模型训练单元,具体用于建立如下所示物性计算模型:
    Figure PCTCN2021098570-appb-100044
    其中,f为所述单分子的物性,m 1i为一级基团中第i种基团的基团数量,Δf 1i为一级基团中第i种基团对物性的贡献值,m 2j为二级基团中第j种基团的基团数量,Δf 2j为二级基团中第j种基团对物性的贡献值;m Nl为N级基团中第l种基团的基团数量,Δf Nl为N级基团中第l种基团对物性的贡献值;a为关联常数;N为大于或等于2的正整数。
  48. 根据权利要求43所述的装置,其特征在于,产品调合单元,具体用于所述单分子的所有基团中确定一级基团、一级基团的基团数量、多级基团和多级基团的基团数量;将构成单分子的所有基团作为一级基团;将同时存在且对同一种物性共同存在贡献的多种基团作为多级基团,将所述多种基团的数量作为所述多级基团的级别。
  49. 根据权利要求48所述的装置,其特征在于,所述产品调合单元,具体用于根据如下物性计算模型计算所述单分子的沸点:
    Figure PCTCN2021098570-appb-100045
    其中,T为所述单分子的沸点,SOL为根据构成所述单分子的每种基团的基团数量转化得到的单分子向量,GROUP 11为根据一级基团对所述沸点的贡献值转化得到的第一贡献值向量,GROUP 12为根据二级基团对所述沸点的贡献值转化得到的第二贡献值向量,GROUP 1N为根据N级基团对沸点的贡献值转化得到的第N贡献值向量,Numh为单分子中除氢原子以外的原子个数,d为第一预设常数、b为第二预设常数、c为第三 预设常数;所述N为大于或等于2的正整数。
  50. 根据权利要求48所述的装置,其特征在于,所述产品调合单元,具体用于根据如下物性计算模型计算所述单分子的密度:
    Figure PCTCN2021098570-appb-100046
    其中,D为所述单分子的密度,SOL为根据构成所述单分子的每种基团的基团数量转化得到的单分子向量,GROUP 21为根据一级基团对所述密度的贡献值转化得到的第N+1贡献值向量,GROUP 22为根据二级基团对所述密度的贡献值转化得到的第N+2贡献值向量,GROUP 2N为根据N级基团对密度的贡献值转化得到的第2N贡献值向量,e为第四预设常数;所述N为大于或等于2的正整数。
  51. 根据权利要求48所述的装置,其特征在于,所述产品调合单元,具体用于根据如下物性计算模型计算所述单分子的辛烷值:
    X=SOL×GROUP 31+SOL×GROUP 32+......+SOL×GROUP 3N+h;
    其中,X为所述单分子的辛烷值,SOL为根据构成所述单分子的每种基团的基团数量转化得到的单分子向量,GROUP 31为根据一级基团对辛烷值的贡献值转化得到的第2N+1贡献值向量,GROUP 32为根据二级基团对辛烷值的贡献值转化得到的第2N+2贡献值向量,GROUP 3N为根据N级基团对辛烷值的贡献值转化得到的第3N贡献值向量;所述N为大于或等于2的正整数;h为第五预设常数。
  52. 根据权利要求36所述的装置,其特征在于,所述混合产品的产品物性,包括:密度、浊点、倾点、苯胺点和辛烷值。
  53. 根据权利要求52所述的装置,其特征在于,所述产品调合单元,具体用于通过如下计算公式计算每组所述混合产品的密度:
    density=Σ(D i×x i_volume);
    其中,density为所述混合产品的密度,D i为第i种所述单分子的密度,x i_volume为第i种所述单分子的第二组分含量。
  54. 根据权利要求52所述的装置,其特征在于,所述产品调合单元,具体用于针对每组混合产品,根据该组混合产品中每种所述单分子的密度和沸点计算得到每种所述单分子的浊点贡献值;根据该组混合产品中所有所述单分子的浊点贡献值和每种单分子的含量,计算该组混合产品的浊点。
  55. 根据权利要求52所述的装置,其特征在于,所述产品调合单元,具体用于针对每组混合产品,根据该组混合产品中每种所述单分子的密度和分子量,计算每种所述单 分子的倾点贡献值;根据该组混合产品中所有所述单分子的倾点贡献值和每种单分子的含量,计算该组混合产品的倾点。
  56. 根据权利要求52所述的装置,其特征在于,所述产品调合单元,具体用于针对每组混合产品,根据该组混合产品中所述单分子的密度和沸点计算得到所述单分子的苯胺点贡献值;根据该组混合产品中所有所述单分子的苯胺点贡献值和每种单分子的含量,计算所述混合产品的苯胺点。
  57. 根据权利要求52所述的装置,其特征在于,所述产品调合单元,具体用于针对每组混合产品,获取该组混合产品中每种所述单分子的辛烷值和每种单分子的含量;通过如下计算公式计算所述混合产品的辛烷值:
    Figure PCTCN2021098570-appb-100047
    Figure PCTCN2021098570-appb-100048
    Figure PCTCN2021098570-appb-100049
    Figure PCTCN2021098570-appb-100050
    Figure PCTCN2021098570-appb-100051
    Figure PCTCN2021098570-appb-100052
    Figure PCTCN2021098570-appb-100053
    其中,所述ON为所述混合产品的辛烷值,HISQFG为分子集合,H为正构烷烃的分子集合,I为异构烷烃的分子集合,S为环烷烃的分子集合,Q为烯烃的分子集合,F为芳香烃的分子集合,G为含氧化合物的分子集合,υ i为所述混合产品中的各个分子的含量;υ H、υ I、υ S、υ Q、υ F、υ G分别为所述混合产品中的正构烷烃的总含量、异构烷烃的总含量、环烷烃的总含量、烯烃的总含量、芳香烃的总含量和含氧化合物的化合物总含量;β i为所述混合产品中的每种分子的回归参数;ON i为所述混合产品中的每种 分子的辛烷值;C H表示正构烷烃与其他分子的交互系数;C I表示异构烷烃与其他分子的交互系数;C S表示环烷烃与其他分子的交互系数;C Q表示烯烃与其他分子的交互系数;C F表示芳香烃与其他分子的交互系数;C G表示含氧类化合物与其他分子的交互系数;
    Figure PCTCN2021098570-appb-100054
    表示正构烷烃与异构烷烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100055
    表示正构烷烃与环烷烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100056
    表示正构烷烃与烯烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100057
    表示正构烷烃与芳香烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100058
    表示正构烷烃与含氧化合物之间的第一常数系数、
    Figure PCTCN2021098570-appb-100059
    表示异构烷烃与环烷烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100060
    表示异构烷烃与烯烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100061
    表示异构烷烃与芳香烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100062
    表示异构烷烃与含氧化合物之间的第一常数系数、
    Figure PCTCN2021098570-appb-100063
    表示环烷烃与烯烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100064
    表示环烷烃与芳香烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100065
    表示环烷烃与含氧化合物之间的第一常数系数、
    Figure PCTCN2021098570-appb-100066
    表示烯烃与芳香烃之间的第一常数系数、
    Figure PCTCN2021098570-appb-100067
    表示烯烃与含氧化合物之间的第一常数系数、
    Figure PCTCN2021098570-appb-100068
    表示芳香烃与含氧化合物之间的第一常数系数、
    Figure PCTCN2021098570-appb-100069
    表示正构烷烃与异构烷烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100070
    表示正构烷烃与环烷烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100071
    表示正构烷烃与烯烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100072
    表示正构烷烃与芳香烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100073
    表示正构烷烃与含氧化合物之间的第二常数系数、
    Figure PCTCN2021098570-appb-100074
    表示异构烷烃与环烷烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100075
    表示异构烷烃与烯烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100076
    表示异构烷烃与芳香烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100077
    表示异构烷烃与含氧化合物之间的第二常数系数、
    Figure PCTCN2021098570-appb-100078
    表示环烷烃与烯烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100079
    表示环烷烃与芳香烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100080
    表示环烷烃与含氧化合物之间的第二常数系数、
    Figure PCTCN2021098570-appb-100081
    表示烯烃与芳香烃之间的第二常数系数、
    Figure PCTCN2021098570-appb-100082
    表示烯烃与含氧化合物之间的第二常数系数、
    Figure PCTCN2021098570-appb-100083
    表示芳香烃与含氧化合物之间的第二常数系数;其中,所述辛烷值包括:研究法辛烷值和马达法辛烷值。
  58. 根据权利要求33所述的装置,其特征在于,所述装置还包括:
    模型训练单元,用于建立产物预测模型;其中,所述产物预测模型,包括:包括多种反应规则的反应规则集合以及反应速率算法;获取样本原料的样本原料信息;利用所述样本原料信息,对所述反应规则集合进行训练,并固定训练完成的所述反应规则集合;利用所述样本原料信息,对所述反应速率算法进行训练,并固定训练完成的所述反应速率算法,得到训练完成的所述产物预测模型。
  59. 根据权利要求58所述的装置,其特征在于,所述样本原料的样本原料信息,包括:所述样本原料的分子组成,所述样本原料中每种分子的分子含量,所述样本原料对应的实际产物的分子组成以及所述实际产物中每种分子的实际含量。
  60. 根据权利要求59所述的装置,其特征在于,所述模型训练单元,具体用于将所述样本原料的分子组成按预设的反应规则集合进行处理,得到所述样本原料的分子组成中每种分子对应的反应路径;根据所述样本原料的分子组成中每种分子对应的反应路径,得到包含所述样本原料、中间产物以及预测产物的装置输出产物的第一分子组成;在所述装置输出产物中,包括:所述样本原料、中间产物以及预测产物;根据所述装置输出产物的第一分子组成与所述实际产物的第二分子组成,计算第一相对偏差;若所述第一相对偏差符合预设条件,则固定所述反应规则集合;若所述第一相对偏差不符合预设条件,则调整所述反应规则集合中的反应规则,根据调整后的反应规则集合,重新计算所述第一相对差值,直至所述第一相对偏差符合预设条件。
  61. 根据权利要求60所述的装置,其特征在于,所述模型训练单元,具体用于获取所述第一分子组成中单分子的种类,构成第一集合;获取所述第二分子组成中单分子的种类,构成第二集合;判断所述第二集合是否为所述第一集合的子集;若所述第二集合不是所述第一集合的子集,则获取预存储的不符合预设条件的相对偏差值作为所述第一相对偏差值;若所述第二集合是所述第一集合的子集,通过如下计算公式计算第一相对偏差:
    Figure PCTCN2021098570-appb-100084
    x 1为所述第一相对偏差,M为所述第一集合,M 1为所述样本原料的分子组成中单分子的种类组成的集合,M 2为所述中间产物的分子组成中单分子的种类组成的集合,M 3为所述第二集合,card表示集合中元素的个数。
  62. 根据权利要求59所述的装置,其特征在于,所述模型训练单元,具体用于根据所述反应速率算法,分别计算所述样本原料的分子组成中每种分子对应的反应路径的反应速率;根据所述样本原料中每种分子的分子含量和所述分子的反应路径对应的反应速率,得到所述样本原料对应的预测产物中每种分子的预测含量;根据所述预测产物中每种分子的预测含量和所述实际产物中每种分子的实际含量,计算第二相对偏差;若所述第二相对偏差符合预设条件,则固定所述反应速率算法;若所述第二相对偏差不符合预设条件,则调整所述反应速率算法中的参数,根据调整后的反应速率算法,重新计算所述第二相对偏差,直至所述第二相对偏差符合预设条件。
  63. 根据权利要求62所述的装置,其特征在于,所述模型训练单元,具体用于根据所述反应速率算法中的反应速率常数,计算每条反应路径的反应速率;
    其中,根据如下计算公式确定所述反应速率常数:
    Figure PCTCN2021098570-appb-100085
    其中,k为所述反应速率常数,k B为玻尔茨曼常数,h为普朗克常数,R为理想气体常数,E为反应路径所处环境的温度值,exp为以自然常数为底的指数函数,ΔS为反应路径对应的反应规则对应的反应前后的熵变,ΔE为反应路径对应的反应规则对应的反应能垒,
    Figure PCTCN2021098570-appb-100086
    催化剂活性因子,P为反应路径所处环境的压力值,α为反应路径对应的反应规则对应的压力影响因子。
  64. 根据权利要求33-63任意一项所述的装置,其特征在于,所述石油加工装置的种类包括:催化裂化装置,延迟焦化装置,渣油加氢装置,加氢裂化装置,柴油加氢改质装置,柴油加氢精制装置,汽油加氢精制装置,催化重整装置和烷基化装置;其中,每种石油加工装置对应一种反应规则集。
  65. 一种分子级装置的实时优化系统,其特征在于,所述分子级装置的实时优化系统包括处理器、存储器;所述处理器用于执行所述存储器中存储的分子级装置的实时优化程序,以实现权利要求1-32中任一项所述的分子级装置的实时优化方法。
  66. 一种计算机可读存储介质,其特征在于,所述计算机可读存储介质存储有一个或者多个程序,所述一个或者多个程序可被一个或者多个处理器执行,以实现权利要求1-32中任一项所述的分子级装置的实时优化方法。
PCT/CN2021/098570 2020-06-12 2021-06-07 一种分子级装置的实时优化方法、装置、系统及存储介质 Ceased WO2021249329A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
EP21822840.1A EP4089680B1 (en) 2020-06-12 2021-06-07 Method, apparatus and system for real-time optimization of molecular-level device, and storage medium
BR112022017194A BR112022017194A2 (pt) 2020-06-12 2021-06-07 Método, aparelho e sistema para otimização em tempo real de dispositivo de nível molecular e meio de armazenamento
US18/047,036 US20230110441A1 (en) 2020-06-12 2022-10-17 Method and System For Real-Time Optimization of Molecular-Level Device, and Storage Medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010533480.3A CN111899793B (zh) 2020-06-12 2020-06-12 一种分子级装置的实时优化方法、装置、系统及存储介质
CN202010533480.3 2020-06-12

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/047,036 Continuation US20230110441A1 (en) 2020-06-12 2022-10-17 Method and System For Real-Time Optimization of Molecular-Level Device, and Storage Medium

Publications (1)

Publication Number Publication Date
WO2021249329A1 true WO2021249329A1 (zh) 2021-12-16

Family

ID=73206275

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/098570 Ceased WO2021249329A1 (zh) 2020-06-12 2021-06-07 一种分子级装置的实时优化方法、装置、系统及存储介质

Country Status (5)

Country Link
US (1) US20230110441A1 (zh)
EP (1) EP4089680B1 (zh)
CN (1) CN111899793B (zh)
BR (1) BR112022017194A2 (zh)
WO (1) WO2021249329A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115831250A (zh) * 2023-02-20 2023-03-21 新疆独山子石油化工有限公司 一种延迟焦化反应模型构建方法及装置、存储介质及设备
CN115938499A (zh) * 2023-02-20 2023-04-07 新疆独山子石油化工有限公司 加氢裂化模型的优化方法、装置、电子设备及存储介质
CN116449691A (zh) * 2022-01-10 2023-07-18 中国石油化工股份有限公司 原料油加工控制方法和装置
CN121034458A (zh) * 2025-10-30 2025-11-28 北京溥络数智科技有限责任公司 一种分子级预测石化生产中任意流股质量的方法及系统

Families Citing this family (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111899793B (zh) * 2020-06-12 2024-04-30 中国石油天然气股份有限公司 一种分子级装置的实时优化方法、装置、系统及存储介质
CN113378332B (zh) * 2021-08-16 2021-11-02 成都数联云算科技有限公司 生产设备组分析方法、装置、电子设备及计算机可读存储介质
WO2023119544A1 (ja) * 2021-12-23 2023-06-29 千代田化工建設株式会社 プログラム、情報処理装置、及び方法
WO2023128828A1 (ru) * 2021-12-29 2023-07-06 Публичное акционерное общество "Газпром нефть" Прогнозирование изменения активности катализатора в установке гидроочистки дизельного топлива
CN116417081A (zh) * 2021-12-31 2023-07-11 中国石油天然气股份有限公司 用于确定油品分子组成数据的解析方法和装置
CN116410772B (zh) * 2021-12-31 2024-04-26 中国石油天然气股份有限公司 一种原油蒸馏切割方法、装置、设备及存储介质
CN116434852A (zh) * 2021-12-31 2023-07-14 中国石油天然气股份有限公司 润滑油生产反应条件确定方法、装置、设备及存储介质和应用
CN116434853A (zh) * 2021-12-31 2023-07-14 中国石油天然气股份有限公司 一种催化裂化过程生产优化方法、装置、设备及存储介质
CN116410771B (zh) * 2021-12-31 2024-03-01 中国石油天然气股份有限公司 一种原油调合优化方法、装置、设备及存储介质
CN114708930B (zh) * 2022-04-19 2025-01-17 泉州装备制造研究所 一种基于层次基团构建的纯组分炼化性质的预测方法
CN115862759B (zh) * 2023-02-20 2023-06-06 新疆独山子石油化工有限公司 一种延迟焦化反应优化方法及装置、存储介质及设备
CN115841851B (zh) * 2023-02-20 2023-06-06 新疆独山子石油化工有限公司 一种加氢裂化分子级反应规则的构建方法和装置
CN116013425B (zh) * 2023-02-20 2023-07-14 新疆独山子石油化工有限公司 一种分子水平炼化过程模型构建方法及装置、存储介质
CN115860267B (zh) * 2023-02-20 2023-06-06 新疆独山子石油化工有限公司 一种炼厂物流价值估算方法及装置、设备及存储介质
CN119875684B (zh) * 2023-10-24 2026-03-10 中国石油化工股份有限公司 渣油加氢操作条件优化方法、装置、电子设备及存储介质
CN118817935B (zh) * 2024-09-14 2025-01-28 苏州博纳化学科技有限公司 水性聚丙烯酸酯和聚氨酯乳液在线反应监测方法及装置
CN119626374A (zh) * 2024-11-20 2025-03-14 香港中文大学(深圳) 量子点手性修饰加工方法、装置、设备、存储介质和程序产品
CN121168188B (zh) * 2025-11-17 2026-02-27 北京溥络数智科技有限责任公司 一种分子级常减压装置的流程模拟优化方法及系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140262957A1 (en) * 2013-03-15 2014-09-18 Exxonmobil Research And Engineering Company Evaluation of distillate composition of a crude
CN109698014A (zh) * 2018-09-25 2019-04-30 中国石油天然气股份有限公司 催化重整产物的预测方法
CN109859805A (zh) * 2019-03-22 2019-06-07 杭州辛孚能源科技有限公司 一种基于分子组成的汽油调和优化方法
CN109949870A (zh) * 2019-03-07 2019-06-28 杭州辛孚能源科技有限公司 一种分子级基础油调和优化方法
CN111899793A (zh) * 2020-06-12 2020-11-06 中国石油天然气股份有限公司 一种分子级装置的实时优化方法、装置、系统及存储介质

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070050154A1 (en) * 2005-09-01 2007-03-01 Albahri Tareq A Method and apparatus for measuring the properties of petroleum fuels by distillation
JP6240091B2 (ja) * 2012-01-13 2017-11-29 アスペン テクノロジー インコーポレイテッド 石油精製のために原油の化学組成をキャラクタリゼーションする方法
CN104484714B (zh) * 2014-11-26 2018-05-11 华东理工大学 一种催化重整装置收率的实时预测方法
FR3030756B1 (fr) * 2014-12-22 2017-01-13 P Plus I Procede de determination de proprietes de coupes petrolieres et des procedes de raffinage pour l’optimisation de raffineries petrolieres.
US10497464B2 (en) * 2015-10-28 2019-12-03 Samsung Electronics Co., Ltd. Method and device for in silico prediction of chemical pathway
CN106444672A (zh) * 2016-10-12 2017-02-22 杭州辛孚能源科技有限公司 针对炼油和石化装置的分子水平的实时优化方法
CN109726844A (zh) * 2017-10-31 2019-05-07 中国石油化工股份有限公司 加氢裂化装置的产品收率自动调控方法、系统和存储设备
CN108279251B (zh) * 2018-01-26 2019-08-02 中国石油大学(北京) 一种石油分子层次分离过程模拟的方法及其装置
CN109960235B (zh) * 2018-11-29 2022-08-26 中国石油化工股份有限公司 基于机理模型的炼化装置实时优化方法和设备
CN110187635B (zh) * 2019-04-10 2022-08-02 浙江中控软件技术有限公司 用于连续重整装置的实时优化方法和设备

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140262957A1 (en) * 2013-03-15 2014-09-18 Exxonmobil Research And Engineering Company Evaluation of distillate composition of a crude
CN109698014A (zh) * 2018-09-25 2019-04-30 中国石油天然气股份有限公司 催化重整产物的预测方法
CN109949870A (zh) * 2019-03-07 2019-06-28 杭州辛孚能源科技有限公司 一种分子级基础油调和优化方法
CN109859805A (zh) * 2019-03-22 2019-06-07 杭州辛孚能源科技有限公司 一种基于分子组成的汽油调和优化方法
CN111899793A (zh) * 2020-06-12 2020-11-06 中国石油天然气股份有限公司 一种分子级装置的实时优化方法、装置、系统及存储介质

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP4089680A4 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116449691A (zh) * 2022-01-10 2023-07-18 中国石油化工股份有限公司 原料油加工控制方法和装置
CN115831250A (zh) * 2023-02-20 2023-03-21 新疆独山子石油化工有限公司 一种延迟焦化反应模型构建方法及装置、存储介质及设备
CN115938499A (zh) * 2023-02-20 2023-04-07 新疆独山子石油化工有限公司 加氢裂化模型的优化方法、装置、电子设备及存储介质
CN121034458A (zh) * 2025-10-30 2025-11-28 北京溥络数智科技有限责任公司 一种分子级预测石化生产中任意流股质量的方法及系统

Also Published As

Publication number Publication date
EP4089680B1 (en) 2025-03-12
BR112022017194A2 (pt) 2022-12-20
EP4089680A4 (en) 2023-08-09
CN111899793B (zh) 2024-04-30
US20230110441A1 (en) 2023-04-13
CN111899793A (zh) 2020-11-06
EP4089680A1 (en) 2022-11-16

Similar Documents

Publication Publication Date Title
WO2021249329A1 (zh) 一种分子级装置的实时优化方法、装置、系统及存储介质
WO2021249330A1 (zh) 一种分子级炼油加工全流程优化方法、装置、系统及存储介质
US10082807B2 (en) Method of preparing a certified target product from a mixture of components by spectral analysis
CN111892960B (zh) 一种汽油调合方法、系统、设备及存储介质
Albahri Structural group contribution method for predicting the octane number of pure hydrocarbon liquids
US8645079B2 (en) Method for measuring the properties of petroleum fuels by distillation
CN111899798B (zh) 一种原油数据管理方法、系统、装置及存储介质
Al Ibrahim et al. Octane prediction from infrared spectroscopic data
CN111899812B (zh) 一种石油加工装置的产物模拟方法、系统、设备和存储介质
CN111892939B (zh) 一种原油蒸馏切割方法、系统、设备及存储介质
CN111892938B (zh) 一种原油分子组成的确定方法、系统、装置及存储介质
CN111899810A (zh) 一种混合物分子组成检测方法、系统、设备和存储介质
Albahri Molecularly explicit characterization model (MECM) for light petroleum fractions
Liu Molecular characterisation and modelling for refining processes
CN116013425B (zh) 一种分子水平炼化过程模型构建方法及装置、存储介质
CN116434847A (zh) 分子级汽油调合配方获取方法、装置、设备及存储介质和应用
CN115860267B (zh) 一种炼厂物流价值估算方法及装置、设备及存储介质
CN116434850A (zh) 分子级汽油调合配方确定方法、设备及存储介质和应用
Stratiev et al. Evaluation of accuracy of literature gasoline blending models to predict octane numbers of gasoline blends
CN116432795A (zh) 润滑油生产加工信息确定方法、装置、设备及存储介质和应用
CN111899794B (zh) 基于b/s架构的炼油装置模拟方法、装置、系统和存储介质
CN116434848A (zh) 燃料油调合配方确定方法、装置、设备及存储介质和应用
CN116434849A (zh) 柴油调合配方确定方法、装置、设备及存储介质和应用
Xiong et al. Well-to-Gate Greenhouse Gas Emissions of Petrochemical Products across the Supply Chain and Sources of Variability in China
Neubauer et al. Surrogates for Liquid–Liquid Extraction

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21822840

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2021822840

Country of ref document: EP

Effective date: 20220809

REG Reference to national code

Ref country code: BR

Ref legal event code: B01A

Ref document number: 112022017194

Country of ref document: BR

ENP Entry into the national phase

Ref document number: 112022017194

Country of ref document: BR

Kind code of ref document: A2

Effective date: 20220826

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 522441195

Country of ref document: SA

WWE Wipo information: entry into national phase

Ref document number: 522441195

Country of ref document: SA

WWE Wipo information: entry into national phase

Ref document number: 522441195

Country of ref document: SA