JPH11265201A - Multivariable process control system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make the setting of the operation restriction condition of a control object process changeable while executing multivariable control without changing the process data base of a basic control base by storing the set value of the operation restriction condition in a data storage means independent of other process data such as an alarm set value. SOLUTION: This control system 1 is provided with the data storage means 12 so as to store restriction condition data used in a multivariable control system separately from the process data base 10. The execution module 13 of the multivariable control system fetches the process data and respective set values of a process 2 from the process data base 10 as data D1013. Also, the restriction condition data D1213 used in the multivariable control system are fetched from the data storage means 12. Based on the fetched data, in the execution module 13, control logic by the multivariable control system is computed and the controlled variable of the process 2 is decided. The arithmetic result is stored in the process data base 10 as the data D1310.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of comparing a control target value, an operation amount, and an operation constraint condition set value of a control amount of a process to be controlled with a measured value of the control amount and an operation amount of the process, and a predicted value thereof. The present invention relates to a multivariable process control system having a function of determining an operation amount of the process based on a result obtained.

[0002]

2. Description of the Related Art In conventional process control, a simple control algorithm is used to determine the manipulated variable of a controlled process due to the difficulty in grasping the characteristics of the controlled process and restrictions on the performance of a control computer. I have been. Especially,
“PID control” that determines a manipulated variable by combining a proportional operation (P), an integral operation (I), and a differential operation (D) with a deviation between a control amount target value and a process response is widely used.

Further, as an important function of the control computer, an operation set point is provided for each of the control amount and the operation amount of the process to monitor the operation. The aforementioned PID control and alarm monitoring
This is "basic control" in process control, and is still used in many plants. On the other hand, in recent process control, the process response prediction calculation is executed using the dynamic characteristic model of the process to be controlled, taking advantage of the high calculation performance of the control computer and the large capacity of the memory, and the response prediction results and control are performed. Comparing the volume target value to determine the optimal manipulated variable,
It has become possible to implement a multivariable control method, and its application to an actual plant is being actively promoted.

[0004] This control method is generally called "model predictive control". In this control method, a deviation between a predicted value of a process control amount and a control amount target value based on the above-described process dynamic characteristic model, and an optimal operation amount that minimizes a change amount of the operation amount as much as possible in each control cycle. This is a method of sequentially calculating.

In this control method, an evaluation function relating to a deviation between a predicted value of a process control amount and a control amount target value and an update amount of an operation amount is derived for each control cycle. This is solved using a numerical optimization technique.
Here, by using the above-described process operation constraint conditions as a search range of a solution of the numerical optimization calculation, the operation amount can be updated in consideration of the behavior range of the control amount and the operation amount. In the above-described basic control based on the PID control method and the alarm monitoring, such an operation amount update is not performed. Therefore, for the control target process, the model predictive control is a multivariable control method that enables more stable control. It can be said that.

However, the above-described model predictive control is a multivariable control method, and it is difficult for the operator to manually intervene. Therefore, it is easy to grasp the operation amount calculation process and to easily perform the manual intervention. In many cases, the above-mentioned basic control is used as a backup or used together. When the characteristics of the controlled process change due to the influence of seasonal changes in outside temperature, etc., the dynamic characteristics model or control system is adjusted.
When it is necessary to individually adjust the response of the control amount or the operation amount, the above-described model predictive control is switched to the basic control method, and the operation of the process is continued. Also, when the above-described model prediction control is applied, the process is first performed after stabilizing the process by the basic control. As described above, the above-described multivariable control method is generally used in combination with the conventional basic control. In such a process control system, process measurement values such as a control amount and an operation amount of a process to be controlled,
The control amount target value and the alarm set value used for the aforementioned basic control are
Store in "process database" with real-time database function. Data items are standardized in the process database so as to store data required for executing the basic control method.

On the other hand, in the model predictive control, which is the above-described multivariable control method, data items such as the above-described driving constraint conditions which are different from the alarm settings of the basic control method are set. in this case,
Although it is necessary to add data items to the above-mentioned process database, it is necessary to reset the data items of the above-mentioned process database every time the input / output relationship of the control target model is changed. Therefore, conventionally, a dedicated computer for model predictive control is installed separately from a control computer for implementing the above-described basic control, and a process database used in model predictive control is set.

In this system configuration, separate process databases are provided for the basic control system and the multivariable control system. However, the same data is stored in the process data to be controlled, so mutual verification of the data is required. Therefore, from the viewpoint of ensuring the reliability of the control system, it is desirable that the process database and the control computer are not separated but shared.

On the other hand, in the process data of the basic control and the model predictive control, the above-described operation constraint is a different data setting item. Unlike the alarm set point in the basic control method, the operation constraint condition is used as a search range when determining an optimal operation amount of the process to be controlled, and thus can be used as an adjustment parameter of a control characteristic.

Therefore, it is necessary to have a function of changing the above-mentioned operation constraint conditions during the execution of the multivariable control without affecting the data items of the process database of the basic control system.

[0011]

SUMMARY OF THE INVENTION An object of the present invention is to mount a conventional basic control system having PID control and an alarm monitoring function and a multivariable control system on the same control computer to ensure reliability. Another object of the present invention is to make it possible to change the operation constraint conditions of the control target process during execution of multivariable control in order to improve control performance without changing the process database of the basic control method.

[0012]

A first aspect of the present invention for achieving the above object is as follows.
The feature of the invention of the present invention is that, in addition to basic process control functions such as alarm monitoring and target value follow-up control of a controlled process, the control amount of the process or its predicted value, the control target value of the process, and the process In a multivariable process control system having a multivariable control function of determining one or more manipulated variables of the process based on the comparison result, the set value of the operating constraint condition is In the case where the multi-variable control is executed by storing the data in a data storage unit independent of other process data, such as an alarm set value and a control target value, the driving constraint condition is used for an operation amount calculation. is there.

According to the first aspect of the present invention, the storage means of the operation constraint condition data used in the model predictive control is made independent of the process database used in the basic control, so that the process database needs to be set for each control method. There is no. As a result, the basic control and the multivariable control can be mounted on the same control computer.

The features of the second invention for achieving the above object are as follows.
Operation management values such as a plurality of control amount target values and alarm set values of the process to be controlled, and used for determining the operation amount of the process,
The set values of the operation constraint conditions related to the manipulated variables and
Displaying a time-varying trend for each variable of the process,
It has a function to enable setting change during execution of multivariable control.

According to the second aspect, during execution of the model predictive control, the process measurement value, the control amount target value, the alarm set value, and the set value of the above-described operation constraint condition are controlled from the process database. , Together with the data display and trend display, the operation constraint setting value change during model predictive control
It can be implemented appropriately.

[0016]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing a functional configuration of a multivariable process control system according to one embodiment of the present invention.

The multivariable process control system 1 of the present invention
Is a process database 10 for storing the controlled variables and manipulated variables of the process 2 to be controlled, and various operating condition data;
An execution module 11 of a conventional basic control method such as D control and alarm monitoring, a data storage unit 12 for storing constraint condition data used for a multivariable control method, and an execution module 13 of a multivariable control method are basic components. Have.

In this embodiment, a control device 3 is provided as a peripheral device for taking in each process measurement signal from the process 2 to be controlled and outputting a control signal to the process 2.
A sequencer, a distributed control device, or the like corresponds to the control device 3. In addition, the process data display device 4
And a data storage and conversion device 5 and a data input device 6 associated therewith.

The process 2 to be controlled according to the present embodiment includes, as basic components, a chemical process device 20, a process detector 21, a process detection device 22 installed in piping, and a control device 23. The process detector 21 installed in the chemical process device 20 has a function of detecting process temperature, pressure, concentration, and the like. On the other hand, the process detector 22 detects a flow rate, a pressure, and the like of a process amount supplied to or released from the process device 20, and the control device 23 is a valve or the like for adjusting the process amount.

The control device 3 of this embodiment captures a signal of each process amount of the process 2 to be controlled and outputs a control signal to each control device. In this embodiment, a case is shown in which S213 and S223 are fetched as process signals and S233 is output as a control signal. The control device 3 converts the process signal fetched from the process 2 into numerical data D311 and
Send to On the other hand, the control system 1 performs alarm monitoring, further performs data processing by PID control or the like, and sets each set value data D113 to the control device.
Send as

The plant data display device 4 of this embodiment displays the process amount of the process 2 and each operation condition data.

The data storage and conversion device 5 stores data D105 to be displayed from the process database 10 provided in the multivariable process control system 1. When the operator changes the display data of the plant data display device 4 using the data input device 6, the data is converted and the data D 510 is transmitted to the process database 10.

Next, a case where a conventional control method such as PID control or alarm monitoring is executed in this embodiment will be described below. In the present embodiment, the control object is to maintain the temperature, which is the internal process amount of the process device 20, in the process 2 at a predetermined value by the flow rate of the heated steam. Here, the steam flow rate is adjusted by the valve 23.

In the above process 2, the temperature is detected by the detector 21. The steam flow is detected by the detector 22. These detection results are taken in the control device 3 as signal levels such as S213 and S223. The control device 3 converts these detection results into numerical data for each control cycle, and transmits the numerical data to the control system 1 as numerical data D311. On the other hand, the control system 1 transmits each set value of the process to the control device 3 as numerical data D113 for each control cycle. The control device 3 converts the data D113 into a signal and outputs it as an opening command signal S233 for the valve 23.

In the control system 1, the device 3
From the execution module 11 of the basic control method
1 is stored in the process database 10. The process database 10 includes a process 2
In addition to the process data D311 described above, operating condition data of the process 2 and data related to the control method are stored. The operation condition data corresponds to, for example, an alarm set point, a temperature set value of the process 2, and the like. The data on the control method corresponds to each gain setting value of the PID control method.

Data D105 selected by the operator from the data stored in the process database 10
Is displayed on the data display device 4 via the data storage and conversion device 5. The data displayed on the data display device 4 is changed by the data input device 6. The updated display result is stored in the process database 10 as data D510.

Next, the execution module 11 of the conventional basic control system will be described. This execution module 11
It has an arithmetic logic, such as a PID control method, conventionally used in process control, and an alarm monitoring function.
The execution module 11 stores, from the process database 10, operating condition data such as process data of process 2 and alarm set values for each control cycle in data D 10.
11 and execute the stored arithmetic logic. The execution result is stored as data D1110 in the process database 10 for each control cycle. In this embodiment, the data D1110 includes the opening command data for the valve 23 in the process 2 obtained by the conventional control method. These data during operation are as described above.
It is displayed on the display device 4.

As described above, the control system 1
When controlling the process 2, all the plant data is managed in the process database 10, and data display and control calculation are executed. The control calculation result is sent to the control device 3 as data D113.
The control device 3 outputs the valve opening command signal S233 to control the opening of the valve 23, thereby controlling the internal temperature of the process device 20 to a predetermined value, which is the control purpose described above.

Next, one embodiment of a multivariable control system of the multivariable process control system 1 of the present invention will be described with reference to FIG.

In the multivariable control method according to the present embodiment, in the operating conditions of the plant, the setting of the operating range of the control amount and the operation amount is treated as a constraint condition in the operation amount calculation.
This is a method in which the operation amount can be determined by directly considering these, and corresponds to a model predictive control method or the like.

In FIG. 1 of the present embodiment, the above-described constraint
2 shows a configuration of the control system 1 for enabling the control system 1 to handle the parameters as control characteristic adjustment parameters. The control system 1 includes a data storage unit 12 for storing constraint condition data used in a multivariable control method, separately from the process database 10 described above. The data stored in the data storage means 12 is displayed on the data display device 4.

The display data D125 is stored in the data storage and conversion device 5 as in the process database 10. When the data displayed on the data display device 4 is changed by the operator, the data D512 after the setting change is transmitted from the data storage and conversion device 5 to the data storage means 12. As a result, it is possible to change the setting of the constraint condition data relating to the multivariable control method without changing the operating conditions used in the conventional control method.

Next, the execution module 1 of the multivariable control method
3 will be described with reference to FIG. The execution module 13 of the multivariable control method, from the process database 10,
The process data of the process 2 and each set value are captured as data D1013. Also, the constraint data D used in the multivariable control method is stored in the data storage means 12.
Capture 1213. On the basis of these fetched data, the execution module 13 calculates a control logic based on a multivariable control method to determine an operation amount of the process 2. The operation result of the execution module 13 is the data D131
0 is stored in the process database 10. Subsequent processing is the same as the above-described conventional control method.

Next, the execution module 13 of the multivariable control system of this embodiment will be described with reference to FIG. The execution module 13 converts the process data D 1013 fetched from the process database 10 into data conversion means 13.
0 and data D12 concerning the constraint condition of the multivariable control method
Thirteen data conversion means 131 and control logic execution means 132 have a basic configuration.

The data conversion means 130 of this embodiment executes data conversion so that the process data D1013 of the process 2 is processed by the control logic execution means 132. Here, in order to execute multivariable control, data normalization and dimensionlessness are executed. Also, if necessary,
Data pre-processing such as weighting of process data, which is not performed by the conventional PID control, is performed. If the control logic 132 employs the model predictive control method, the data conversion means 130 predicts and calculates changes in the process 2 using process data. As described above, the data conversion result for the multivariable control method is sent to the control logic execution unit 132 as data D1302.

The data conversion means 131 of this embodiment performs data conversion of the constraint data used in the multivariable control method. In this case, similar to the above-described data conversion means 130,
Performs normalization and dimensionless constraint data. The data conversion result is stored in the execution unit 1 as data D1312.
32.

In the control logic execution means 132 of this embodiment, the process data D1302 and the constraint data D1
312 is fetched and an operation based on the multivariable control method is executed. Here, when the model predictive control is adopted, the process data D1302 and the constraint condition data D1312
Then, an evaluation function for determining the operation amount is constructed, and the optimum operation amount for the process 2 is determined by numerical optimization.
The calculation result is sent to the process database 10 as data D1310. After being stored in the process database 10, the same processing as in the above-described conventional control method is performed, and the opening degree command signal S233 for the valve 23 is processed.
Is generated.

As described above, the constraint data used in the multivariable control method is stored in the process database 10 by another means, so that when the process 2 is operated by the multivariable control method, the constraint data is stored. Even when the operating condition data such as the conditions are changed and then the operation is switched to the conventional control method, the operating condition data in the process database 10 has not been changed, so that there is no need to adjust again.

Next, the data display and the setting change in the data display device 4 during the operation of the process 2 by the multivariable process control system according to the present embodiment will be described with reference to FIGS.
Will be described.

FIG. 3 shows a display example of a data display screen in the case where the operation status of the process 2 is monitored on the data display device 4 of the present embodiment. In this figure, a monitor screen 40a for process data registered as a control amount and a monitor screen 40b for process data registered as an operation amount in the process 2 are shown. The correspondence between the process amount and the detection point shown in FIG.
The following description is based on the assumption that the "temperature" and the detector 22 correspond to "111 flow rate".

The monitoring screens 40a and 40b can be displayed on the data display device 4 either in a case where either one is displayed or in a case where the display contents are switched by an operator. 3 and 4 of this embodiment, only the operation condition data portion of the process 2 is displayed, and the control amount target value, the process amount at the current time, and the like are omitted.

In FIG. 3 of this embodiment, the operating condition data used in the conventional control method is displayed at 410a and 410b. This data corresponds to an alarm set point or a threshold value for canceling automatic control, and setting is not usually changed during automatic control of a plant. This operation range display data 4
10a and 410b are stored in the process database 10 and are used in the above-described conventional control method as determination data of PID control release input / release.

On the other hand, the display data 411a, 411 of FIG.
b corresponds to the display content of the plant constraint condition data used in the above-described multivariable control method. These display data are stored in the above-mentioned data storage means 12, and are used during the control logic operation when the above-described multivariable control method is executed. In the present embodiment, the operation data is displayed in a table format for each process amount for batch monitoring of the process 2.

The set values of the above constraint data are changed by the data input device 6 of FIG. FIG. 4 shows an embodiment in the case where the “control range” display data is changed.

In this embodiment, it is assumed that the data input device 6 has a data indicating function for specifying data on the data display device 4 and a function for inputting numerical data. Usually, these functions include a mouse and a trackball. The latter corresponds to the keyboard.

FIG. 4 shows a case where the operator instructs the "control range" data of "111 flow rate" on the manipulated variable monitoring screen 40c. The arrow 41 is a display data instruction function. When an instruction is given to a data display portion whose setting can be changed, the designated data can be changed in setting, and the display of the data input screen and the data display specification are changed. In the present embodiment, a situation is shown in which the data instruction function 41 is instructing control range data of “111 flow rate”. In this state, the operator uses the data input device 6 to input a setting change value.

FIG. 4 shows the monitor screen after changing the set value in the manipulated variable monitor screen 40d. MV monitoring screen 40
In d, the control range data of “111 flow rate” is “7.0”
From "6.0" to "6.0". As described above, the data whose setting has been changed on the data display device 4 is stored in the constraint condition data storage unit 12 used for the multivariable control method via the data storage and conversion device 5. For this reason, even if the control range is changed, the control characteristics are changed only when the above-described multivariable control method is executed, regardless of the conventional control method.

Next, a change in the process amount of the process 2 when the constraint data is changed as shown in FIG. 4 of this embodiment will be described with reference to FIG.

FIG. 5 of the present embodiment shows the process data change of the process 2 when the multivariable control method is executed.
It is an example in which a trend is displayed. On this trend monitoring screen 42, the multi-variable control method
The process trend is displayed by the operator's selection. In FIG. 5, "101 temperature" and "111 flow rate"
Is displayed. In each trend display, the vertical axis represents the engineering unit of process data, and the horizontal axis represents time.

The details of each trend display data on the monitoring screen 42 will be described below. The “101 temperature” trend display 420 relates to data registered with a control amount. Therefore, the control amount target value is displayed as the trend D420d. In addition, the control range data 411a of the control amount monitoring screen 40a is displayed as D420a and D420b. Also, the 101 temperature trend is similarly displayed at D420c.

Next, the "111 flow rate" trend 421 will be described. In this trend display, the control range data 411b of the manipulated variable monitoring screen 40b is displayed as D421
a and D421b. Further, the 111 flow rate trend is indicated by D421c in the same manner as in the case of the control amount.

Subsequently, as shown in FIG. 4 of this embodiment,
The following describes a change in the process 2 and a trend display when the constraint condition data is changed during execution of the multivariable control method.

In FIG. 4 of this embodiment, the operation amount monitoring screen 40
In d, the control range data of the 111 flow rate is set to “7.0”.
The case where the setting has been changed from “6.0” to “6.0” is shown. FIG.
Shows the trend change of the process 2 when the setting change time is executed around 0:40. As a result, the data D421a becomes 6.0 after time 0:40.
Is displayed as a trend.

Further, when executing the multivariable control method,
The control range of the 111 flow rate is changed between 6.0 and 5.0. In the multivariable control execution module 13 of FIG. 1 of the present embodiment, in the case where the 111 flow rate is in the range of 6.0 to 5.0, the temperature of the 101 follows the predetermined control amount target value. Determine the opening command.

As a result, in the trend 420 of FIG. 5, the 111 flow rate falls between 6.0 and 5.0 after time 0:40, and the 101 temperature follows the target value trend D420d. Can be changed as follows.

[0057]

According to the present invention, a multivariable control system using the operating conditions of a plant as a constraint is adopted, and apart from the process database used in the conventional basic control system,
By providing the constraint condition data storage means to be considered in the multivariable control method, the multivariable control method can be executed without changing the condition setting of the conventional control method. Also,
The setting of the constraint data can be changed during the execution of the multivariable control, and the operating state and control characteristics of the process can be adjusted.

[Brief description of the drawings]

FIG. 1 is a diagram showing a schematic configuration of a multivariable process control system according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a detailed configuration of an execution module of the multivariable control system in FIG. 1;

FIG. 3 is a view showing a part of display contents of a process amount monitoring screen displayed on the data display device of FIG. 1;

FIG. 4 is a view showing display contents when the setting of constraint data used in the multivariable control method is changed on the process amount monitoring screen displayed on the data display device of FIG. 1;

FIG. 5 is a view showing an example of a trend change when constraint condition data is changed on a process trend display screen displayed on the data display device of FIG. 1;

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Multi-variable process control system, 2 ... Control target process, 3 ... Control device, 4 ... Data display device, 5 ... Data storage and conversion device, 6 ... Data input device, 10 ... Process database, 11 ... Execution of basic control system module,
12 ... Constraint condition data storage means, 13 ... Multivariable control method execution module, 20 ... Plant equipment, 21, 22 ... Process amount detector, 23 ... Valve.

Claims (2)

[Claims] 1. A control amount of a process or a predicted value thereof, a control target value of the process, and a control target value of the process, together with basic process control functions such as alarm monitoring and target value tracking control of a process to be controlled. In a multivariable process control system having a multivariable control function for comparing one or more operation amounts of the process based on the comparison result with a driving constraint condition, the set value of the driving constraint condition is:
An alarm set value and a control target value are stored in a data storage unit independent of other process data, and when the multivariable control is executed, the driving constraint condition is used for an operation amount calculation. And a multivariable process control system. 2. An operation management value such as a control amount target value or an alarm set value of a plurality of control target processes, and a set value of an operation amount used for determining an operation amount of the process and operation constraint conditions related to the control amount. 2. The multivariable process control system according to claim 1, further comprising a function of displaying a time change trend for each variable of the process and enabling a setting to be changed during operation.