JP4324960B2 - Multivariable predictive control system - Google Patents

Description

  The present invention relates to a multivariable predictive control system having a plurality of control processes and a plurality of multivariable predictive control controllers corresponding to these control processes.

  Prior art documents related to the multivariable predictive control system include the following.

URL: http://www.yokogawa.co.jp/EXASOFT/08_Download/download.htm Multi-variable model predictive control system “EXASMOC” (EXASMOC is a registered trademark)

  The document “TI36J06D10-01” that can be downloaded from Non-Patent Document 1 has a technical disclosure on multivariable control and model predictive control. In a control process that has multiple variables that have complex correlations with each other, various constraints can be imposed by relying on the operator's judgment and experience to select variables and the amount of operation for each variable to perform an optimal operation. It is difficult to protect and achieve optimal operation.

The multivariable controller disclosed in Non-Patent Document 1 inputs multivariables having a correlation and realizes the following functions.
(1) Observe the constraints of input / output variables (2) Set the control variable to the steady state with the optimum value (3) Set the manipulated variable to the steady state with the optimum value (4) Prevent excessive fluctuation of the manipulated variable (5) ) Optimum plant operation

  The model predictive control controller disclosed in Non-Patent Document 1 has a process model therein, and predicts an optimum operation amount such that the response of the actual process matches the response of the process model. The controller uses this information to control the manipulated variable so that the control variable finally becomes the target value.

  FIG. 3 is a functional block diagram showing an example of a control system configuration of a plurality of processes in which a multivariable predictive control controller having such multivariable control and model predictive control functions is introduced. Reference numerals 11, 21, and 31 denote correlated processes, and reference numerals 41 and 51 also denote correlated processes.

  12, 22, and 32 are multivariable predictive control controllers corresponding to the processes 11, 21, and 31 at 1: 1. Similarly, 42 and 52 are multivariable predictive control controllers corresponding to the processes 41 and 51 at 1: 1.

  Reference numerals 13, 23, and 33 denote operation / monitoring means corresponding to the multivariable predictive control controllers 12, 22, and 32 at 1: 1. Similarly, 43 and 53 are operation / monitoring means corresponding to the multivariable predictive control controllers 42 and 52 at 1: 1.

  In the conventional system configuration, the operation / monitoring means in the introduction of the multivariable predictive control controller to a plurality of correlated control processes is connected as a host device of the multivariable predictive control controller per control process, The multivariable predictive control controller was individually handled and optimization was performed for each process.

  A multivariable predictive controller has a lot of input / output data from the control process, and there is a correlation that affects the input / output data of other multivariable predictive control controllers in the control process. obtain.

  In a conventional system configuration using individual operation / monitoring means, an operator needs to access a plurality of operation / monitoring means for a process having a correlation. Therefore, it is complicated to operate and monitor the entire correlated process, which is an obstacle to optimum operation.

  Furthermore, since the number of input / output data handled by the multi-variable predictive control controller over a plurality of processes becomes enormous, an environment for centrally managing these data cannot be achieved with the conventional configuration.

  Therefore, the problem to be solved by the present invention is to realize a multivariable predictive control system capable of centrally managing input / output data of a plurality of multivariable predictive control systems corresponding to a plurality of control processes. is there.

In order to achieve such an object, the configuration of the present invention is as follows.
(1) In a multivariable predictive control system having a plurality of control processes and a plurality of multivariable predictive control controllers corresponding to these control processes,
A plurality of multivariable predictive control controllers having a plurality of control parameters and performing multivariable predictive control for each control process by operating in independent control cycles ;
Both the governing said plurality of multivariable predictive controller, the main controller providing the running optimization operation of Constrained relates multivariable predictive controller, the setting range of optimal control parameters for the multivariable predictive controller When,
Operation / monitoring means for controlling and monitoring the main controller;
A multivariable predictive control system characterized by comprising:

(2) The operation / monitoring unit displays an operation / monitoring screen in which the main controller, the multivariable predictive control controller, and the control process are displayed in a hierarchy with the self as the highest level Multivariable predictive control system.
(3) The multivariable predictive control system according to (1) or (2), wherein the operation / monitoring screen opens a detailed screen of a corresponding item by clicking an item displayed in a tree shape.

( 4 ) The multivariable predictive control system according to any one of (1) to (3), wherein the control parameter is at least one of an operation variable, a control variable, a disturbance variable, and a process output.

As is apparent from the above description, the present invention has the following effects.
(1) A multi-variable predictive control controller corresponding to a plurality of correlated control processes is controlled by a main controller, and a process having a correlation is provided by adopting a hierarchical structure in which operation / monitoring means is provided above the main controller. It is possible to control and monitor the entire system, and to provide an optimal operation environment in which complicated operations are eliminated.
(2) It is possible to provide an environment in which a large amount of input / output data handled by a multivariable predictive control controller over a plurality of processes is centrally managed by a higher-level operation / monitoring means.

  Hereinafter, the present invention will be described in detail with reference to the drawings. FIG. 1 is a functional block diagram showing an embodiment of a multivariable predictive control system to which the present invention is applied. The same elements as those in the conventional system described with reference to FIG. Hereinafter, the characteristic part of the present invention will be described.

  In FIG. 1, reference numeral 100 denotes a multivariable prediction control unit that supervises the control processes 11, 21, 31 having correlation. Reference numeral 200 denotes a multivariable prediction control unit that controls the control processes 41 and 51 having correlation.

  In the multivariable predictive control unit 100, 101 is a sub-controller that executes multivariable predictive control for the control process 11, 102 is a subcontroller that executes multivariable predictive control for the control process 21, and 103 is multivariable predictive control for the control process 31. Is a sub-controller that executes

  In the multivariable prediction control unit 100, reference numeral 104 denotes a main controller, which supervises the sub-controllers 101 to 103 and optimizes the evaluation function provided in each sub-controller.

  Similarly, in the multivariable prediction control unit 200, 201 is a sub-controller that executes multivariable prediction control for the control process 41, and 202 is a subcontroller that executes multivariable prediction control for the control process 51.

In the multivariable predictive control unit 200, 203 is a main controller, which supervises the sub-controllers 201 and 202, and optimizes the evaluation function provided in each sub-controller.
The sub-controllers 101, 102, 103, 201, and 202 correspond to the multivariable prediction control unit controller in the conventional example of FIG.

  Reference numeral 300 denotes an operation / monitoring unit located at the highest level, and performs operation / monitoring by supervising the main controller 104 of the multivariable prediction control unit 100 and the main controller 203 of the multivariable prediction control unit 200.

  Thus, in the present invention, the level (A) of the operation / monitoring means 300 located at the highest level, the level (B) of the main controllers 104 and 203, the levels (C) of the sub-controllers 101 to 103 and 201 and 202, and the lowest level. It is characterized by a system configuration in which the levels (D) of the control processes 11, 21, 31, 41, 51 located in FIG.

  The operation / monitoring means 300 located at the highest level displays an operation / monitoring screen in which the main controller, the sub-controllers, and the control processes are shown in a hierarchy with the self as the highest. In this operation / monitoring screen, a detailed screen of the corresponding item is opened by clicking an item displayed in a tree shape.

  FIG. 2 shows an example of a hierarchical display, in which the level (A) of the highest level operation / monitoring means to the level (D) of the lowest level control process are hierarchically displayed. Specific display items of the lowest-level control process are an operation variable, a control variable, a disturbance variable, and a process output. Clicking on this displays detailed data belonging to this item.

  The operator can centrally operate and monitor all the data of the correlated processes by the hierarchical screen of the operation / monitoring means 300, and can access a plurality of operation / monitoring screens. The operation is canceled.

  The sub-controller operates in an independent control cycle, performs multivariable predictive control for each control process, and has a plurality of control parameters. The control parameter is at least one of the aforementioned operation variable, control variable, disturbance variable, and process output.

  The main controller executes a controllability check and an evaluation function optimization calculation in consideration of the constraint conditions related to the sub-controller, and gives an optimal control parameter setting range to the sub-controller.

  In the embodiment described above in FIG. 1, the operation / monitoring unit 300 controls the main controllers 104 and 203 in the two groups of the groups 100 and 200. However, a configuration in which other groups are also controlled is possible. A configuration in which one group of main controllers is integrated may be employed.

It is a functional block diagram which shows one Embodiment of the multivariable predictive control system to which this invention is applied. It is an example of a display of the operation and monitoring screen displayed hierarchically by this invention. It is a functional block diagram which shows an example of the conventional multivariable predictive control system.

Explanation of symbols

11, 21, 31, 41, 51 Control process 100, 200 Multivariable predictive control controller 101, 102, 103, 201, 202 Sub controller 104, 203 Main controller 300 Operation / monitoring means

Claims (4)

In a multivariable predictive control system having a plurality of control processes and a plurality of multivariable predictive control controllers corresponding to these control processes,
A plurality of multivariable predictive control controllers having a plurality of control parameters and performing multivariable predictive control for each control process by operating in independent control cycles ;
Both the governing said plurality of multivariable predictive controller, the main controller providing the running optimization operation of Constrained relates multivariable predictive controller, the setting range of optimal control parameters for the multivariable predictive controller When,
Operation / monitoring means for controlling and monitoring the main controller;
A multivariable predictive control system characterized by comprising: 2. The multivariable prediction according to claim 1, wherein the operation / monitoring unit displays the operation / monitoring screen in which the main controller, the multivariable predictive control controller, and the control process are hierarchized with self as the highest level. Control system. The multi-variable predictive control system according to claim 1 or 2, wherein the operation / monitoring screen opens a detailed screen of a corresponding item by clicking an item displayed in a tree shape. The multivariable predictive control system according to any one of claims 1 to 3, wherein the control parameter is at least one of an operation variable, a control variable, a disturbance variable, and a process output.

Images