JPS59168502A - Weight setting system for calculation of optimum control value - Google Patents

Abstract

PURPOSE:To calculate the optimum control value that is accordant with the actual object process characteristics by controlling the weight at a working field in response to the dynamic responsiveness of the object process. CONSTITUTION:The dynamic response is calculated for each part of an object process 1 which received a command from a control setting/changing part 2 which gives a process dynamic response to decide the weight. Then the time series data is stored in a memory part 3, and the control deviation value and the manipulated variable which are related to decide the weight are shown in a diagram in terms of time series out of the data stored in the part 3. At the same time the present set weight is displayed on the screen of a CRT4. Looking at this display, a setting/changing part 5 increases (decreases) the weight of the control deviation value or the manipulated variable that has an extraordinary large (small) amount of variation. Then an optimum control value calculating part 6 performs a matrix operation to minimize the control index. A control part 7 calculates the optimum manipulated variable from the optimum control value.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a control method when a target process has a plurality of controlled variables and a plurality of manipulated variables for controlling these variables.

Conventionally, in this type of control device, a control method has been configured for each controlled variable by preselecting a controlled variable in which the control effect of each manipulated variable mainly appears. However, in this type of control system, the control method is designed ignoring the influence of other manipulated variables on the controlled variable, so if the control characteristics deteriorate or in extreme cases, it may become unstable. There were some drawbacks that could lead to this. On the other hand, from multiple control variables, the following 1
There is a method in which two control indexes J are set and control is performed so that these are minimized.

J=L(q+(xIs XI)2+(12(X2S
Xz)"+-)d+X4 Xz...... is the control amount, and XI'S X2S is their control setting. q+
qz... is the control coordination weight for each control amount.

In this case, there is a method of setting these weight values in advance, assuming the target process characteristics, calculating the optimal adjustment value, and using it for control, but in reality, the target process characteristics are the assumed characteristics. The optimum adjustment value may not be obtained due to the difference between the two values.

This invention is a method that enables calculation of optimal adjustment values that match the actual characteristics of the target process by observing the dynamic response characteristics of the target process and adjusting these weights in the field. It has a structure that allows easy human operation for on-site weight adjustment.

The figure shows the functional configuration of an embodiment of the invention. 1 is the target process part, 2 is the control setting change part that gives the process dynamic response to determine the weight, 3 is the dynamic response accumulation part, and 4 is the weight of the accumulated dynamic response. The display part of the response related to the setting and the display part of the current setting weight, 5 is the weight change part, 6 is the calculation part of the optimal adjustment value for the given weight, and 7 is the optimal adjustment found by this. This is the part that calculates the amount of operation for the target process based on the value and outputs the actual amount of operation to control the target process.

2 includes a function of selecting the type of control variable to which a setting change is desired from among a plurality of control variables of the target process, and a function of issuing an actual setting change command.

3 receives the setting change command from 2, measures the dynamic response of each part of the target process, and stores the time-series data in the storage section. 4 also displays the current setting weight on the accumulation screen 3 at the same time. A person looks at the screen No. 4, and increases the weight of the corresponding amount with an unreasonably large variation among the control deviation and manipulated variable, and decreases the weight of the corresponding amount with an unreasonably small variation.

These weight setting changes are performed in step 5. 6 is 5
A well-known matrix operation is performed to minimize the control index J for the weight given by , and an optimal adjustment value is calculated.

7 multiplies the optimum adjustment values of the dynamic basic quantities of the target process to obtain the optimum operating quantities, and outputs these to the target process.

The above weight adjustment process is repeated until the process converges. What is obtained as a result is the weight for that process operating state, and is the five optimal adjustment values.

According to the present invention, since the weights are set in the field, it is possible to set the weights that allow calculation of the optimal adjustment value that suits the target process characteristics. Therefore, the direction of weight modification can be physically understood, and weight setting operations by humans are facilitated.

[Brief explanation of drawings]

- is a functional configuration diagram of the present invention. 1: Target process, 2: Control setting change part, 3: Dynamic response accumulation part, 4: Display part, 5: Weight setting change part, 6: Optimum adjustment value calculation part, 7: Optimal manipulated variable output part.

Claims (1)

[Claims] In a control method that creates one control index for multiple control variables of a controlled process by considering the weight of each control coordination, and coordinately controls multiple manipulated variables to minimize this control index, A part that changes the control settings, a part that accumulates the dynamic response of the target process at that time, a part that displays time difference sequence data of control deviation amount and manipulated variable on the screen from the accumulated data, a part that changes the weight settings, and the weight. A weight setting method for calculating optimal adjustment values for control, which includes an optimal adjustment value calculation part for the optimum adjustment value, and a part that calculates the optimal operation amount based on the optimal adjustment value and outputs it to the target process.