JPH0273402A - Plant preventive maintenance method and preventive maintenance equipment - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a method and apparatus for performing preventive maintenance of a plant composed of a large number of devices, and is particularly applicable to preventive maintenance of a nuclear power plant. This relates to methods and equipment suitable for this purpose.

[Conventional technology]

BACKGROUND ART Conventionally, the maintenance and maintenance of various equipment constituting a plant has been carried out as periodic inspections, such as periodic inspections and replacement of worn parts. However, the data for determining the frequency of inspection and replacement was determined based on estimated lifespan values through life acceleration tests, empirical values, etc., and was not necessarily the best for each component.

Regarding such maintenance and inspection, please refer to Japanese Unexamined Patent Publication No. 62-27647.
The technique described in No. 01 is publicly known.

[Problem to be solved by the invention]

In order to guarantee the reliability and safety of plant operation, it is desirable to increase the frequency of periodic inspections, increase the number of inspection items, and standardize inspection standards.

However, from an economic point of view, it is desirable to avoid unnecessary inspections and replacements as much as possible.

In order to satisfy these conflicting demands to the maximum extent possible, it is not appropriate to carry out uniform periodic inspections of the large number of devices that make up the plant; instead, it is necessary to take into account the characteristics and history of each individual device. It won't happen.

For example, during regular inspections.

(a) For equipment that has recently had frequent minor failures, the number of inspection items should be increased and inspection standards should be particularly strict;
For equipment that has been replaced entirely with new equipment, inspection standards can often be relaxed.

Conventionally, judgments regarding each piece of equipment have been made manually by plant management engineers by looking at the equipment's resume or recent operating data.

However, manually examining a large number of pieces of equipment individually and making appropriate decisions not only requires a great deal of effort and time, but also requires plant management engineers to have extremely advanced technical knowledge. Experience required.

Moreover, there is no guarantee that there will be no human error.

The present invention has been made in view of the above-mentioned circumstances, and it is possible to develop the history of each component device without requiring special skill or a great deal of effort, and without the possibility of making a human error.

It is an object of the present invention to provide a method that can automatically plan preventive maintenance for the entire plant, taking into account the current situation, and a device that can also automatically plan preventive maintenance.

[Means to solve the problem]

The basic principle of the preventive maintenance method of the present invention created to achieve the above object can be summarized as follows.

(i) Group the large number of devices that make up the plant.

There may be a group unit consisting of one device, and
A large number of devices of the same type may be made into one group.

(n) Each group is provided with a device-specific subsystem that has the function of receiving, solving, determining, and storing data issued by one device.

(iii) Provide a host system that integrates multiple model-specific subsystems.

As a specific configuration for practically applying the above principle, the preventive maintenance method of the present invention is such that the device-specific subsystem has a function of estimating the remaining life of the device, and the host system , create a maintenance plan for the entire plant based on information from each equipment subsystem.

Further, the device of the present invention created to carry out the above method of the invention is provided with a device-specific subsystem having a function of diagnosing the health of each component device and estimating the remaining life, and A host system is provided that has the function of creating a maintenance plan for the entire plant based on the output information of the system.

[Effect]

When the preventive maintenance method described above is carried out using the preventive maintenance device with the above configuration, the device-specific subsystem estimates the health and remaining life based on the device-specific data and sends the results to the host system. Therefore, the host system can comprehensively judge the input information and create a preventive maintenance plan for the entire plant, taking into account the technical characteristics (history, operating status, etc.) of each component. I can do it.

When implementing the method of the present invention, by establishing a knowledge base and updating or adding to the knowledge base, it is possible to take into account the past history of the device and to obtain information based on the latest information. Preventive maintenance can be performed.

If you have the above knowledge information such as information interpretation, condition diagnosis, remaining life prediction, and maintenance inspection method determination, you will be able to create a complete preventive maintenance plan. It is valid. Even if one part of this knowledge is used, a corresponding effect can be expected.

Further, when implementing the present invention, if the host system is provided with an information storage means and an inference calculation means for processing the stored information, a more sophisticated preventive maintenance plan can be created.

By applying automatic calculation means and automatic storage means in this way,
The creation of preventive maintenance plans is automated, which not only achieves labor savings, but also eliminates the need for the personal skill and experience of management engineers, eliminating the possibility of human error. .

〔Example〕

An embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a conceptual diagram of a system according to an embodiment of the present invention. FIG. 2 is a system processing flow diagram of one embodiment of the present invention, showing the processing contents of the device-specific subsystem.

The device-specific subsystems 4-1 to 4-3 shown in FIG.
There are i pieces in total, each of which receives device data 6 as information from one or more devices, and also receives additional knowledge ii7.

The host system 2 includes the i device-specific subsystems 4.
It is connected to subsystems 4-1 to 4-1 by communication lines, and centrally manages these subsystems 4-1 to 4□ while exchanging information.

Reference numeral 1 in the figure indicates input information of the host system 2, which specifically includes maintenance slip information, accident information, plant data, and the like. This host system 2 mainly has the function of determining the next periodic inspection process. In the next process, preventive maintenance information for each device immediately before periodic inspection will be sent from the device-specific subsystem 4.

For example, the inspection priority order is revised and determined. Also, the next process is corrected one after another starting from the corrected next process.

The device-specific subsystem 4 provides a subsystem for each important device in the plant, such as pumps and valves, so that information common to pumps, for example, can be stored as a database.
Along with data sharing, it diagnoses the condition, predicts the remaining lifespan, determines inspection methods, and outputs preventive maintenance information for each device. Therefore, it is possible to perform simultaneous distributed processing of diagnosis for each device. In this way, the distributed processing of the subsystem 4 for each device and the function of the host system 2 is effective in speeding up the system and improving reliability. The following explanation uses a pump as an example.・Repair slip (
MRF) information, accident information such as equipment troubles, and plant data such as process data are input to the host system 2 as input information 1, and the database management control unit 3 identifies the information and updates the old data to create common plant data. is converted into a database in the host system 2, and the pump individual data is transferred to the equipment subsystem 4, and managed as the equipment side database 18 by the database management control section 5 in the subsystem. Data discrimination in the database management control units 3 and 5 in both the equipment and the host system is done by attaching a label to each piece of data to determine whether it is plant-common data or equipment-specific data.

Furthermore, in order to be able to determine which processing is required for the inference calculation section of both the equipment subsystem and the host system, the data is labeled with a keyword in advance to indicate whether it is related to a work process or lifespan when inputting the data. It has become Therefore, the database management control units 3 and 5 can determine the label of data, and can perform data exchange between the subsystem and host system and output appropriate data in response to data input requests from each calculation unit.

In the equipment subsystem 4, knowledge such as remaining life prediction, maintenance and inspection methods, and condition diagnosis is stored in a knowledge base 17 in the form of rules for processing preventive maintenance information of pumps.

Figure 5 is a specific example of knowledge being made into rules. As shown in Figure 5, rules of thumb are organized as production rules.
Factual data and codes of conduct are organized using frames.

Figure 2 shows the device-specific subsystems 4-1 to 4 shown in Figure 1.
- is a flow diagram showing the contents of each. In the knowledge base 17 shown in the figure, input/output of knowledge data to the inference calculation section 11 is managed by a knowledge pace management section 16. Further, the knowledge in the knowledge base 17 can be input from the knowledge base management unit 16 to the knowledge base construction support tool 8, and the knowledge can be updated and expanded to change the knowledge base 17.

The specific functions of Knowledge Base Construction Support Tool 8 are (
i) PTA event editing function i Display the FTA graph, add, delete, move, copy elements, change element names, set and change connection logic, and add reliability (confidence, weighting coefficient). (il) Display screen scrolling function, (correct) knowledge expression definition function: Defines knowledge in a specified range to be expressed in a frame or production rule. (tv
) Terminology exchange function: When diagnostic knowledge is partially input using a pre-registered technical term dictionary, candidate terms to be used uniformly are expressed on the CRT. (v) Knowledge registration function; registers the edited knowledge in the knowledge base 17.

FIGS. 6 and 7 are specific examples of acquiring and modifying knowledge using the knowledge base construction support tool. The step of changing the FTA section is performed as shown in FIG. 6 using the mechanism (i) described above. Note that the connection relationship between single elements is automatically regarded as an OR element.

- Power tube 7 The causal elements shown in Figure 7 are the causes of abnormal events such as cracks and corrosion, and the observed events are findings based on observation, such as "pump shaft vibration exceeds 20% of the specified value." For editing these causal elements and observation elements, display the entire knowledge in the format shown in Figure 7, and
Enter 2.

If the entire knowledge cannot be displayed on the screen, the screen can be moved using the scroll function.

When editing the FTA section and the causal element section, the terminology conversion function allows unified technical terminology.

For example, reactor water/reactor water is selected as reactor water level.

In the inference calculation unit 11 (FIG. 2), the database 18
Based on the data in the knowledge base 17, inferences are made in each program of the equipment remaining life prediction processing 132, the equipment state diagnosis processing 14, and the equipment inspection method determination processing 15. The equipment remaining life prediction process predicts the remaining life of the pump, and the equipment status diagnosis process determines the necessity of maintenance.

The equipment inspection method determination process determines the contents of pump inspection and maintenance items. The input information interpretation program 12 performs natural language interface processing with the operator, and can also specify the driving order and driving program for each of the inference processes at the request of the operator.

The preventive maintenance information for each device that has been variously inferred by the inference calculation section 11 is transferred to and from the host system 2 via the database management control sections of both the sub and host systems. Further, the device-specific subsystem 4 outputs device-specific preventive maintenance information 10.

FIG. 3 shows the processing contents of the host system and is a processing flow diagram of one embodiment of the present invention.

The preventive maintenance information transferred from the device-specific system 4 to the host system 2 is temporarily stored in the database management control unit 3 of the host system 2.

It is used by the periodic inspection plan processing program 26 of the inference calculation unit 20 as necessary. On the other hand, the accident information stored in the database of the host system 2 is transmitted to the improvement item processing program 22 of the inference processing section 20 via the database management control section 3.
It is determined whether the item is related to equipment improvement or work improvement, and if it is a work improvement item, it is processed into an improvement proposal suitable for the plant in the work improvement processing program 24, and is sent to the periodic inspection plan processing program 2 as work improvement information.
6.

In addition, in the case of equipment improvement items, equipment improvement processing program 23
and create improvement proposals suitable for the plant in question.

The periodic inspection process processing program 27 calls the results, schedule, and plant operation information of the periodic inspection process from the database 19, and automatically updates the data regarding the results and schedule of the periodic inspection process. Further, in response to a request from the periodic inspection plan processing program 26, the next periodic inspection process information is output to the periodic inspection plan processing program 26.

The pending issue processing program 25 executes pending information processing such as periodic inspection remaining items for the plan 1 and repair slip information.

For example, a list is created and output in order of processing urgency.

A periodic inspection plan processing program into which preventive maintenance information for each device temporarily stored in the database management control unit 3, work improvement items from the work improvement processing program 24, and next periodic inspection process information from the periodic inspection process processing program 21 are input. 26 executes the inspection and maintenance target equipment, inspection and maintenance contents, work improvement items, and optimal work schedule that can be performed within the next periodic inspection process period by inference processing.

As described above, the plant comprehensive preventive maintenance information 9, such as equipment improvement proposals, issue management, next periodic inspection plans, and private inspection items, processed by each program of the inference calculation unit 20 is output to the outside via the man-machine interface 21. be done. The man-machine interface 21 inputs and outputs various types of information and inputs and outputs information related to system control through screen displays. Further, in the device-specific subsystem 4, the input information processing program 12 and the man-machine interface are linked, and natural language interface processing can be performed.

In this embodiment, the knowledge content of the knowledge base 17 is input to the knowledge base construction support tool 8 via the knowledge base management unit 16, and after updating and expanding the knowledge content, it is returned to the knowledge base 17. I can do it.

In this way, the knowledge content stored in the knowledge base 17 (for example, knowledge about the status of a piece of equipment, knowledge for predicting the remaining life, knowledge for determining maintenance and inspection methods, etc.) is always kept up to date. It can be maintained as

In this way, by updating and adding to the knowledge regarding the status of the device, the best possible consideration, prediction, and decision can be made taking into account the history of the device.

〔Effect of the invention〕

According to the method of the present invention, it is possible to determine whether or not maintenance and inspection of equipment is necessary.

By automatically evaluating remaining life prediction and proposing the next periodic inspection items and contents for the entire plant, a rational periodic inspection plan is automatically created to reduce and prevent equipment failures and defects. be done.

This improves the operational reliability of the plant and significantly reduces management costs.

Moreover, according to the apparatus of the present invention, the above-mentioned method of the invention can be easily and reliably carried out to exhibit its effects, and individual You can obtain a preventive maintenance plan for the entire plant that takes into account the history and current status of the equipment.

[Brief explanation of the drawing]

Fig. 1 is a system conceptual diagram of an embodiment of the device of the present invention;
The figure is a processing flow diagram showing the processing contents of the device-specific subsystem of FIG. FIG. 3 is a processing flow diagram showing the processing contents of the host system of FIG. 1. FIG. 4 is a system function flow diagram in an embodiment of the present invention. FIG. 5 is also an explanatory diagram of the abnormality diagnosis rule. Fig. 6 is also an explanatory diagram of the correction operation of the diagnostic PTA, Fig. 7
The figure is an explanatory diagram of correction of cause elements and observed event elements other than PTA. 1... Input information, 2... Host system, 3...
Database management control unit, 4...device-specific subsystem. 8...Knowledge base construction support tool, 9...Comprehensive preventive maintenance information, 10...Preventive maintenance information by device, 13...
- Equipment condition diagnosis processing program, 14... Remaining life prediction processing program, 15... Maintenance inspection method determination processing program, 23... Equipment improvement processing program, 25...
Pending matter processing program, 26... Periodic inspection plan processing program.

Claims (1)

[Claims] 1. In a preventive maintenance method for a plant constituted by a large number of devices, (a) the above-mentioned large number of devices are grouped into a single device or a plurality of devices; (b) each of the above-mentioned devices is A device-specific subsystem having a function of estimating the remaining life of devices belonging to each group is provided for at least a plurality of groups, and (c) a means of communication is provided between the plurality of device-specific subsystems. (d) the host system creates a plant maintenance plan based on the estimated remaining life of a plurality of equipment-specific subsystems that are in communication with the host system; Plant preventive maintenance methods. 2. A knowledge base is provided in at least one of the device-specific subsystem and the host system, and this knowledge base is updated or added to through interaction with a user, according to claim 1. Plant preventive maintenance methods. 3. A database is provided in the device-specific subsystem to store at least one of device usage information, history information, and maintenance information, and to automatically perform input/output with the host system based on this stored information. Manage and establish a knowledge base to create rules and accumulate knowledge for interpreting information for preventive maintenance, diagnosing conditions, predicting remaining life, and determining maintenance and inspection methods. The preventive maintenance method for a plant according to claim 1 or 2, characterized in that: 4. A database management means is provided in the host system to store information for preventive maintenance processed by each equipment subsystem, and to store at least one of accident information, operation information, and equipment information for the entire plant. In addition to accumulating one piece of information, an inference calculation means is provided to automatically output at least one of periodic inspection contents, periodic inspection plan, equipment improvement proposal, and pending issue processing information based on the above accumulated information. A preventive maintenance method for a plant according to claim 1 or 3, characterized in that the method is characterized in that: 5. In a device that performs preventive maintenance of a plant composed of a large number of devices, (a) a signal emitted from the device is input to each one or more of the devices to check the health of the device; (b) The output of the plurality of device-specific subsystems is inputted to the entire plant. A preventive maintenance device for a plant, characterized by having a host system equipped with a function of creating a maintenance plan. 6. A knowledge base is provided in at least one of the device-specific subsystem and the host system, and the knowledge base is configured to be able to be updated and/or added to at least one of the above through interaction with the user. The preventive maintenance device for a plant according to claim 1. 7. In the device-specific subsystem, (a) provide a database capable of storing at least any of device usage information, history information, status information, and maintenance information, and (b) provide data input/output to the above database. A database management control unit that has a control function and a function to update data stored in the database, and (c) knowledge to interpret information input for each device and diagnose the condition of the device. and a knowledge base having a function of accumulating at least one of knowledge for predicting remaining life and knowledge for determining maintenance and inspection methods, and (d) provided from the database and knowledge base. An inference calculation unit is provided that has functions for diagnosing the condition of the equipment, predicting its remaining life, and determining maintenance and inspection methods based on the information, and (e) output signals from the database, the knowledge base, and the inference calculation unit; and a man-machine interface for processing input signals of external information.
The plant preventive maintenance device according to claim 5. 8. The plant preventive maintenance system according to claim 7, wherein the knowledge base includes a knowledge base construction support tool that supports expansion and modification of the knowledge base. 9. The above-mentioned host system has (a) a function to temporarily store information processed by the equipment-specific subsystem and a function to accumulate technical information regarding the plant, and controls the input and output of the host system. (b) an inference calculation unit that creates and determines at least one of periodic inspection details, periodic inspection plans, equipment improvement proposals, and pending processing information based on the accumulated data; (c) a man-machine interface for processing output signals from the database, output signals from the inference calculation unit, and external signals input to the host system. , A preventive maintenance device for a plant according to claim 5 or 7.