JP2000214903A - Parameter adjustment system for controller using perturbation - Google Patents

Abstract

PROBLEM TO BE SOLVED: To automatically adjust the parameter for a controller optimizing an evaluation function by using a perturbation. SOLUTION: The value of each parameter is corrected so as to reduce the value of an evaluation function by utilizing the difference of an evaluation function between when perturbations are simultaneously applied to all parameters and when they are not. In such a case, the evaluation function J (p) is an evaluation index representing whether the entire control system performs desired operation, and is generally given with the error between a plant output and its target value. It is also a function of a parameter for a controller. The initial numeric value (p) of a certain parameter vector is used for the controller to operate the control system and the value J (p) of the evaluation function is calculated. Next, a perturbation vector (c) is added to the parameter (p), and an evaluation function J (p+c) is found by using the parameter value of p+c. The updation quantity Δp of the parameter (p) is calculated according to the values of the two evaluation functions and the value of the parameter (p) is updated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an automatic adjustment method using a perturbation of a parameter of a controller in a control system.

[0002]

2. Description of the Related Art In a control system, it is desired to adjust a parameter p such as a gain in a controller so that an output y of a plant to be controlled matches a target value r as quickly as possible.

Conventionally, appropriate parameters of the controller are generally determined based on various theories by analyzing a control system. For this reason, this determination requires much labor for analysis and the like, and when the characteristics of the control target change, readjustment is often required.

FIG. 4 shows a configuration diagram of a PID control system generally used. Here, the KP included in the controller,
By appropriately adjusting parameters called gains of TI and TD, the output y of the plant is made to coincide with the target value r as quickly as possible. These parameters are often adjusted by a person based on empirical rules, and are time-consuming.

[0005]

In the conventional method, the adjustment and determination of parameters have many parts based on plant analysis and human experience, and are time-consuming. An object of the present invention is to provide a method for automatically obtaining a controller parameter that optimizes a certain evaluation function by using perturbation.

[0006]

The invention according to claim 1 is
A method of automatically adjusting parameters such as gain included in the controller to an appropriate value, and a value of an evaluation function representing an evaluation of the entire control system when a small perturbation is given to all parameters simultaneously, It is configured by a parameter adjustment method of a controller characterized by a function of adjusting these parameters to appropriate values based on the difference from the value of the evaluation function when no perturbation is given.

According to a second aspect of the present invention, there is provided a method for automatically adjusting a parameter such as a gain included in a controller to an appropriate value, wherein a perturbation is superimposed when the parameter is adjusted. It is constituted by the parameter adjustment method of the controller according to the first aspect.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a flowchart showing an embodiment of a system for automatically adjusting a parameter of a controller using a simultaneous perturbation type optimization method to which the present invention is applied. FIG. 2 is a flowchart showing an embodiment of a method for automatically adjusting a parameter of a controller using a simultaneous time difference perturbation optimization method according to a second embodiment.

In FIG. 1 and FIG. 2, an evaluation function J
(P) is an evaluation index indicating whether or not the entire control system is performing a desired operation, and is generally given as an error between the output of the plant and its target value. It is also a function of the parameters of the controller.

In FIG. 1, a control system is operated by using an initial value p of a certain parameter vector as a controller, and a value J (p) of an evaluation function is obtained. Next, the perturbation vector c is added to the parameter p, and the evaluation function J (p + c) is obtained using the parameter value of p + c. From the values of these two evaluation functions, the update amount Δp of the parameter p is calculated, and the parameter p
Update the value of. The updating algorithm is shown in equations (1) and (2).

[0011]

(Equation 1)

Here, Δpi represents the ith element of the update vector Δp of the parameter vector, α is a positive number, and ci is the ith element of the perturbation vector c. Further, t represents time.

In the parameter adjusting mechanism, by sequentially updating the parameters of the controller based on the above-mentioned algorithm, the value of the parameter is adjusted so as to indicate the value of the optimum evaluation function.

The algorithm for updating the parameter value is not limited to the equations (1) and (2), but may be the following equations (3) and (4).

[0015]

(Equation 2)

Here, each coefficient is the same as that in the equations (1) and (2).

[0017]

As described above, according to the present invention, the adjustment of the parameters of the controller in the control system is automatically adjusted to an appropriate value regardless of the analysis of the characteristics of the controlled object or the experience or trial of a person. Can be adjusted.

[Brief description of the drawings]

FIG. 1 is a flowchart illustrating an embodiment of the present invention.

FIG. 2 is a flowchart illustrating an embodiment of the present invention.

FIG. 3 is a diagram showing an example of a step response when the present invention is used in a control system.

FIG. 4 is a block diagram of a general PID control system.

[Explanation of symbols]

 y Output of controlled object r Target value of output c Perturbation vector p Parameter vector included in controller J (p) Evaluation function t Time α Positive number

Claims (2)

[Claims] 1. A method for automatically adjusting a parameter such as a gain included in a controller to an appropriate value, and represents an evaluation of an entire control system when a small perturbation is given to all of the parameters at the same time. A parameter adjustment method for a controller characterized by a function of adjusting these parameters to appropriate values based on a difference between a value of an evaluation function and a value of an evaluation function when no perturbation is given. 2. A method for automatically adjusting a parameter such as a gain included in a controller to an appropriate value, wherein a perturbation is superimposed when the parameter is adjusted.
3. The parameter adjustment method of the controller according to 1.