JPH02207301A - Process controller - Google Patents

Abstract

PURPOSE:To effectively control an internal state change at a relatively long cycle due to the fluctuation of a process operating state by successively proofreading an assumed model formula in order to determine an intermediate variable target value. CONSTITUTION:An assumed model proofreading means 14 compares a control object analysis value inputted from an analyzing means 12 with a control object assumed value 1 preserved in an assumed value preserving means 11 by taking a time delay portion into consideration, and successively proofreads the assumed model formulas of a control object assuming means 10 and an intermediate variable target value determining means 13. A comparing means 15 compares an intermediate variable value (k) from an intermediate variable value operating means 9 with an intermediate variable target value (o) from the intermediate variable target value determining means 13, and an operation edge output means 16 determines an operation edge target value (s) to be an output value to a process 1. Thus the internal stage change at the relatively large cycle due to the fluctuation of the process operating state is effectively controlled.

Description

[Detailed Description of the Invention] [Object of the Invention] (Industrial Application Field) The present invention relates to a process control device used when the characteristics of a process are complex and there are many state variables. The present invention relates to a process control device that allows a controlled object to follow a target value while sequentially adapting.

(Prior art) In recent years, in interference processes in which manipulated variables and internal state variables are intricately intertwined, and in addition, in processes where the controlled object cannot be measured online and has dead time that can be known through manual analysis, it is becoming increasingly difficult to It has become an important issue to maintain a preset target value and reduce variations.

Conventionally, two types of devices have been used as control devices for this type of process. That is, one of them is, as shown in Fig. 3, the sampled value a sampled periodically from the process 1 is analyzed by the analyzer 2, and the analyzed value of the controlled object from this analyzer 2 is converted into the controlled object target value. C
is compared by the comparator 3, and based on the comparison result, the operating end target value d, which is the output value to the process, is set by the operating end output device g.
i''4 and outputs it to process 1 for control.The other is to determine the state of input value e from process 1 using state detection device 5, as shown in FIG. Based on the process internal state value f from the state detection device 5, the intermediate variable value calculation device 6 calculates an intermediate variable value that serves as an index for process control, and calculates the intermediate variable value g from the intermediate variable value calculation device 6. is compared with the intermediate variable target value by the comparator 7, and based on the comparison result, the operating end target value d, which is the output value to the process, is determined by the operating end output device 4, and this is output to the process 1. It is controlled by

However, in a process control device like the one shown in Figure 3,
Since the information used for feedback includes wasted time in process 1 and wasted time due to manual analysis,
There is a problem in that control responsiveness deteriorates significantly in response to disturbances with a short period or disturbances with a period shorter than the sampling period for manual analysis.

In addition, the process control device shown in Figure 4 enables online feedback control, which solves the problems of dead time and sampling period mentioned above, but the relationship between intermediate variables and controlled objects remains constant. Since it is based on this assumption, there is a problem that it is not effective when the relationship changes from time to time due to fluctuations in the operating state of Process 1.

(Problems to be Solved by the Invention) As described above, in conventional process control devices, the control response to disturbances with a period shorter than the sampling period is significantly reduced, and relatively due to fluctuations in the operating state of the process. There is a problem in that it is difficult to effectively control long-period changes in internal conditions.

An object of the present invention is to enable effective control even in response to changes in internal conditions over a relatively long period due to fluctuations in process operating conditions, without reducing control responsiveness to disturbances with a period shorter than the sampling period. Our goal is to provide highly reliable process control equipment.

[Structure of the Invention] (Means for Solving the Problems) In order to achieve the above object, the process control device of the present invention includes a state detection means for detecting the operating state of the process online, and an online control system for detecting the operating state of the process. Based on the detected information, an intermediate variable value calculation means that calculates an intermediate variable value that is an index for process control, and based on the online detection information from the state detection means and the intermediate variable value from the intermediate variable value calculation means,
A controlled object estimating means for estimating a controlled object of process control using a predetermined estimation model formula, an analysis means for analyzing sampling values periodically sampled from the process, online detection information from a state detecting means and a controlled object an intermediate variable target value determining means for determining an intermediate variable target value necessary for making the controlled object follow the controlled object target value based on the target value using the same estimation model formula as the controlled object estimating means; Every time a controlled object analysis value is input from the means, the controlled object estimation means and intermediate variables are compared by comparing the controlled object analysis value and the controlled object estimated value from the controlled object estimation means, taking into account the time delay. The estimated model calibration means sequentially calibrates the estimated model formula of the target value determining means, and the process and an operating terminal output means for determining an operating terminal target value which is an output value to the operating terminal.

(Function) Therefore, in the process control device of the present invention, the operating state of the process is detected online, the intermediate variable value that is an index for process control is determined from the operating state of the process, and the intermediate variable value and the intermediate variable value are determined. Since the process is controlled by determining the operating terminal target value from the comparison result with the variable target value, the control responsiveness does not deteriorate with respect to a disturbance having a period shorter than the sampling period. In addition, an estimation model equation for estimating the controlled object of process control from the operating state of the process and intermediate variable values, and a method for making the controlled object follow the controlled object target value from the operating state of the process and the controlled object target value are also provided. The estimated model equations for determining the required intermediate variable target values are sequentially calibrated using the analyzed values of the controlled object, so that they can be calibrated against relatively long-period changes in internal conditions due to fluctuations in process operating conditions. It also becomes possible to effectively control.

(Example) Hereinafter, an example of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing an example of the configuration of a process control device according to the present invention. That is, as shown in the figure, the process control device of this embodiment includes a state detection means 8, an intermediate variable value calculation means 9, a controlled object estimation means 10, an estimated value storage means 11, an analysis means 12, and an intermediate variable value calculation means 9. Target value determining means 1
3, estimated model calibration means 14, comparison means 15, and operating end output means 16.

Here, the state detection means 8 receives the input value l from the process 1.
In other words, the operating status of the process is detected online. The intermediate variable value calculating means 9 calculates an intermediate variable value k, which is an index for process control, based on the process internal state value j from the state detecting means 8. Also,
The controlled object estimating means 10 estimates the controlled object of the process control using a predetermined estimation model formula based on the process internal state value j from the state detecting means 8 and the intermediate variable value from the intermediate variable value calculating means 9. It is estimated. Further, the estimated value storage means 11 stores the controlled object estimated value l from the controlled object estimation means 10 in chronological order (along with the estimated time).

On the other hand, the analysis means 12 analyzes the sampling values m periodically sampled from the process 1.

Further, the intermediate variable target value determining means 13 includes the state detecting means 8
Based on the process internal state value j and the controlled object target value n from This determines the value 0. Furthermore, each time the controlled object analysis value p is input from the analysis means 12, the estimated model calibration means 14 calculates the controlled object analysis value p and the controlled object estimated value l stored in the estimated value storage means 11 over time. The estimated model equations of the controlled object estimating means 1o and the intermediate variable target value determining means 13 are sequentially calibrated (the calibration value is set to q) by comparing them in consideration of the delay.

Further, the comparison means 15, in the intermediate variable value calculating means 9, compares the opening degree value with the intermediate variable target value 0 from the intermediate variable target value determining means 13. Furthermore, the operating end output means 16 determines the operating end target value S, which is the output value to the process 1, based on the comparison output from the comparing means 15.

In this case, the estimated model equation in the controlled object estimating means 10 is determined in advance by a mathematical method such as a regression analysis method using, for example, actual measured values obtained from a dynamic characteristic test of an actual process (1). The equation and the estimated model equation in the intermediate variable target value determining means 13 are also the same as (2
) formulas are used respectively.

l Hina ni 10 ni 2・ni + ΣGI Tatami j1...(1)
n-Kl + 2 ・o+ΣG1-31 Therefore, o= in-(Kt+ΣGi-jl)1/K 2
...(2) Here, Kl,
2, Gl is each constant of the estimated model formula, and the estimated model formula is calibrated by calibrating + among these constants.

Next, the operation of the process control device configured as above will be explained.

In FIG. 1, the state detection means 8 detects the state of the input value from the process 1, and the detected process internal state value j is input to the intermediate variable value calculation means 9. Then, the intermediate variable value calculating means 9 calculates an intermediate variable value k, which is an index of process control, from the process internal state value j, and
This intermediate variable value is input to the comparison means 15. Also,
In the intermediate variable target value determination means 13, based on the process internal state value j from the state detection means 8 and the controlled object target value n,
An intermediate variable target value 0 necessary for causing the controlled object to follow the controlled object target value n is determined using the estimated model equation (2), and this intermediate variable target value O is also input to the comparison means 15. As a result, the comparing means 15 uses the intermediate variable value from the intermediate variable value calculating means 9 as the intermediate variable target value determining means 1.
3 is compared with the intermediate variable target value O. Then, the operating end output means 16 outputs the comparison output r from the comparison means 15.
Based on this, the operating end target value S, which is the output value to the process 1, is determined, and by outputting this operating end target value S to the process 1, feedback control for the intermediate variable target value 0 is performed.

On the other hand, the intermediate variable value calculated by the intermediate variable value calculating means 9 is also input to the controlled object estimating means 10. As a result, the controlled object estimation means 10 uses the estimation model equation (1) based on the process internal state value j from the state detection means 8 and the intermediate variable value from the intermediate variable value calculation means 9. The control target for process control is estimated. The controlled object estimated value l obtained by the controlled object estimation means 10 is sequentially stored in the estimated value storage means 11 together with the estimated time.

On the other hand, the analysis means 12 analyzes the sampling values m periodically sampled from the process 1, and this controlled object analysis value p is input to the estimated model calibration means 14.

Then, in the estimated model calibration means 14, every time the controlled object analysis value p is input from the analysis means 12, the controlled object analysis value p and the controlled object estimated value l stored in the estimated value storage means 11 are adjusted with a time delay. A calibration value q is obtained by comparing the calculated values in consideration of the difference, and this calibration value q is inputted to the controlled object estimating means 10 and the intermediate variable target value determining means 13, respectively. As a result, the estimated model equations of the controlled object estimating means 10 and the intermediate variable target value determining means 13 are successively calibrated using the calibration value q.

Note that FIG. 2 is a diagram showing a temporal transition between the intermediate variable target value 0 and the intermediate variable calculated value in FIG. 1.

As described above, in the process control device of this embodiment, the intermediate variable value k, which serves as an index for process control, is determined from the operating state j of the process detected online, and the operation is performed from the intermediate variable target value 0 to this intermediate variable value. Since the final target value S is determined and the process 1 is controlled, the above-mentioned third
It is possible to solve the problem of the process control device shown in the figure, in which control responsiveness is significantly reduced to disturbances having a period shorter than the sampling period, and to perform good control. In addition, an estimation model equation for estimating the controlled object of process control from the operating state j of process 1 and the intermediate variable value, and the operating state of the process, j and the controlled object target value n
The estimated model formula for determining the intermediate variable target value 0 necessary to make the controlled object follow the controlled object target value n is
Since each analysis value m of the controlled object is sequentially calibrated, it is possible to avoid relatively long period changes in internal conditions due to fluctuations in the operating state of process 1, which is the problem with the process control device shown in Figure 4 mentioned above. This solves the problem of not being able to effectively control the process, making it possible to perform extremely reliable process control.

〔Effect of the invention〕

As explained above, according to the present invention, an intermediate variable value serving as an index for process control is obtained from the operating state of the process detected online, and an operating terminal target value is determined from this intermediate variable value and intermediate variable target value. Estimated model formula for controlling the process and estimating the controlled object of process control from the operating state of the process and intermediate variable values, and tracking the controlled object to the controlled object target value from the operating state of the process and the controlled object target value Since the estimated model equations used to determine the target values of intermediate variables required to achieve the desired values are sequentially calibrated using the analyzed values of the controlled object, control responsiveness to disturbances with periods shorter than the sampling period decreases. without doing,
Moreover, it is possible to provide an extremely reliable process control device that can effectively control even relatively long-period changes in internal conditions due to fluctuations in process operating conditions.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of a process control device according to the present invention, FIG. 2 is a diagram showing temporal changes in intermediate variable target values and intermediate variable calculated values in the same embodiment, and FIGS. Each figure is a block diagram showing an example of a conventional process control device. l... Process, 8... State detection means, 9... Intermediate variable value calculation means, 10... Controlled object estimation means, 11
... Estimated value storage means, 12... Analysis means, 13...
- Intermediate variable target value determining means, 14... Estimated model calibration means, 15... Comparing means, 16... Operating end output means. Figure 2

Claims (1)

[Claims] A process control device used when a process has complex characteristics and many state variables includes a state detection means for detecting the operating state of the process online, and a state detection means based on online detection information from the state detection means. an intermediate variable value calculating means for calculating an intermediate variable value serving as an index for process control; and a predetermined estimation model based on the online detection information from the state detecting means and the intermediate variable value from the intermediate variable value calculating means. Controlled object estimating means for estimating a controlled object of process control using a formula; Analysis means for analyzing sampling values periodically sampled from the process; Online detection information and controlled object target value from the state detecting means. intermediate variable target value determining means for determining an intermediate variable target value necessary for making the controlled object follow the controlled object target value using the same estimation model formula as the controlled object estimating means; Each time a controlled object analysis value is input from the analysis means, the controlled object estimation means compares the controlled object analysis value with the controlled object estimated value from the controlled object estimation means, taking into account the time delay. and estimated model calibration means for sequentially calibrating the estimated model formula of the intermediate variable target value determining means, and a comparison result between the intermediate variable value from the intermediate variable value calculating means and the intermediate variable target value from the intermediate variable target value determining means. 1. A process control device comprising: operating end output means for determining an operating end target value, which is an output value to the process, based on the above.