JPH0231204A - Control device - Google Patents

Abstract

PURPOSE:To correctly reflect the result of improvement for the response of a closed loop by instructing the start for observation of a waveform when a fixed useless time passed after the change of an arithmetic parameter. CONSTITUTION:The dynamic characteristics of a control subject 1 are changed with the change of the target control value and various types of disturbance. A PI control means 2 performs a PI operation based on a deviation signal DV obtained between the process value PV given from the subject 1 and the target control value SV and outputs the obtained manipulated variable MV to the subject 1. A parameter tuning means 3 observes the waveform of the response signal of the PV or the DV and tunes the PI arithmetic parameter of the means 2 so that the observed waveform is used as the target of response. A monitor means 4 monitors whether the absolute value of the DV is kept within a range of the prescribed value or not. A waveform observation instruction means 5 detects the parameter changed by the means 3 and instructs the observation of waveform of the response signal after a fixed time if the parameter is changed.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a regulating device used for process control;
More specifically, the present invention relates to a self-tuning adjustment device that observes the waveform of a signal from a process and automatically adjusts at least proportional (P) and integral (I) calculation parameters to optimal values.

(Prior Art) In process regulators used for feedback control, PI calculation parameters have traditionally been set manually based on the long-standing knowledge and experience of process operators or instrumentation engineers. This is the current situation.

However, when using manual methods, under conditions such as process start-up, unexpected disturbances, or systems with nonlinear gain characteristics, the process may experience operational disturbances, either temporally or steadily. This may have caused economic losses.

Therefore, an adjustment device that self-tunes the PI calculation parameters has been proposed.

FIG. 5 is based on US Pat. No. 4,602,326
1-245203) is a configuration block diagram of a self-tuning adjustment device.

This regulating device has a process variable PV from process 1;
First comparator 20 that calculates the deviation of the control target 1sv
a detection means 21 which observes the behavior of the process jiPV and detects the characteristics of the pattern; and a second detection means 21 which is coupled to this detection means 21 and which compares the detected characteristic value of the pattern with a desired value of the pattern set in advance. Comparator 22;
The adjusting means 2 is coupled to the second comparator 22 and outputs the manipulated variable MV so that the process variable PV matches the control target value SV.

The detection means 21 detects the behavior (response) of the process quantity PV, and when the magnitude of this behavior exceeds a predetermined dead band, the detection means 21 observes and analyzes the peak value of the response signal, the time between peaks, etc. However, the overshoot amount and damping are determined, and based on these values, the PI calculation parameters are manipulated and changed. The self-tuning regulator of
This method is superior in that it does not have any influence on the process because it does not forcefully apply an identification signal that causes disturbance to the process.

However, by observing the peak (extreme value) of the response signal, P
Since the I calculation parameters are changed and waveform observation is started immediately after that, the results of improved closed-loop response due to the changed calculation parameters (in principle, the response signal appears after the minimum process dead time has elapsed). ) was not reflected correctly, and as a result, incorrect calculation parameters were calculated and set.

The present invention has been made in view of these points, and
The purpose is to realize a self-tuning adjustment device that can always set and change accurate calculation parameters.

(Means for Solving the Problem) FIG. 1 is a block diagram showing the basic configuration of the present invention. In FIG. It is assumed that the dynamic characteristics change due to changes in values, various disturbances, etc. 2 is a PI control means that performs at least P,■ calculation on the deviation signal DV between the process amount P■ from the controlled object 1 and the control target value S■, and outputs the obtained manipulated variable MV to the controlled object 1; is 10
Observe the response signal waveform of the deviation amount P■ or deviation signal I) V, and adjust P so that the observed waveform becomes the desired response target.
Parameter tuning means for tuning the PI calculation parameters of the I control means; 4 a monitoring means for monitoring whether the absolute value of the surface difference signal DV falls within a predetermined value range or whether this absolute value increases beyond a predetermined value; , 5 detects a parameter change by the parameter tuning means 3,
This is a waveform RiII instruction means that instructs to observe the response signal waveform after a certain period of time when a parameter is changed.

(Function) The waveform 1f11 measurement instruction means 5 detects a parameter change by the parameter tuning means 3, and when there is a parameter change, it determines whether the absolute value of the deviation signal DV is within a predetermined value range or if this absolute value is within a predetermined value range. Provided that the value has increased beyond the value of , an instruction is given to start waveform observation after a certain dead time has elapsed after changing the calculation parameters. This allows waveform observation to be performed on the response signal that reflects the current calculation parameters set in the Pl control means 2.

(Example) Embodiments of the present invention will be described in detail below with reference to the drawings.

FIG. 2 is a configuration block diagram showing an embodiment of the present invention. In FIG. 0, 6 is a controller which inputs the process amount PV and control target value S
Output V to the controlled object 1.

In this controller, 61 is the process amount PV, the control item i
A multiplexer which sequentially selects and inputs the value Sv and various analog signals ei, 62 a comparator which receives the signal selected by the multiplexer 61 as one input, and 63 a microprocessor to which the signal from the comparator 62 is input.

64 is a D/A converter that converts the digital signal from the microprocessor 63 into an analog signal, and 65 is 1)/A.
This is a sample and hold circuit that holds the signal from the converter 64 at a predetermined timing, and its output is sent to the controlled object 1 as the operation signal MV. 66 is an I10 board, 67 is a RAM that stores various data, 68 is a main operating program executed by the microprocessor, and P in FIG.
A ROM that stores programs for functioning as the I control means 2, parameter tuning means 3, monitoring means 4, and waveform observation instruction means 5, and 69 is a user ROM that stores programs created by the user according to the process, for example.
, 70 are display keyboards which are coupled to the microprocessor 63 via a data bus BS.

The main operations of the apparatus configured as described above will be explained below.

FIG. 3 is a flowchart showing an example of the operation performed by the microprocessor.

The microprocessor 63 first stores the system ROM 6.
8, the process amount Pv, control target value SV, and other analog signals e1 applied to the multiplexer 61 are sequentially selected and extracted, and these signals are sent to the comparator 62, the microprocessor 63, and the microprocessor 63. The A/D conversion loop formed by the /A converter 64 converts each into a digital signal (step 1).

Next, a deviation signal D between the process amount PV and the control target value S■
Step 2: Determine whether V is within the predetermined set value range DB (deviation set value for self-tuning startup)
), when DV>DB, move to the next step 3. In step 3, the waveform of the deviation signal DV is observed in accordance with the system program functioning as the parameter tuning means 3, and the pattern of the observed waveform is analyzed. Furthermore, when the pattern analysis is completed, the evaluation index of the observed waveform is calculated (step 4.5). A method such as that proposed in No. 53494 is used.

After completing the calculation of the evaluation index of the observed waveform, the parameter tuning means 3 compares these evaluation indexes with each data of the predetermined optimal response model, calculates PI calculation parameters, and sends this to the PI control means 2. Setting (tuning) (step 6), based on the evaluation index,
As a method for calculating the PI calculation parameters, for example, the method disclosed in the aforementioned Japanese Patent Application No. 62-53494 is used.

The PI control means 2 performs control calculations using the PI calculation parameters set and changed by the parameter tuning means 3, and outputs the calculation results to the controlled object 1 as the operation signal MV (step 7.8).

Thereafter, the monitoring means 4 monitors whether the absolute value of the interval signal DV has reached a predetermined value, for example (1/2) DB, or whether the absolute value of the deviation signal DV has increased beyond a predetermined value, for example DB. (Step 9.10).

Subsequently, the waveform a measurement instruction means 5 performs the PI measurement in step 6.
Determine whether or not the calculation parameters have been changed, and if there is a change, perform timer processing (Step 111.1)
2.13), as a result, after a certain dead time has elapsed, the process returns to step 1 and the waveform ff1j in step 3 is
Instruct measurement.

Figure 4 shows, for example, I) with respect to changes in the process amount P■.
3 is a diagram showing how calculation parameters (proportional gain) are sequentially changed by the parameter tuning means 3. FIG.

In (a), time t when the process amount PV starts to fluctuate
1, the first waveform observation is started. When the analysis of the waveform pattern of the process amount Pv is completed at time t2, parameter tuning is performed based on the analysis result, and P
As shown in (b), the calculation parameter is changed from Pl, which had been set up to that point, to 22. Here, at the point X where the process amount Pv crosses 0, it becomes Yes in step 9,
After the P calculation parameter is changed to P2, a certain dead time TL (this dead time can be a dead time measured in advance by process identification or a value set as a parameter by the user). At the elapsed time t3, the answer is Yes in step 13, and the next waveform observation is started. The waveform observation here reflects the result of changing the calculation parameter to P2.

This waveform observation ends at time t4, and based on this analysis result, the next parameter tuning is performed, and the PIi4
3! , the parameter is changed from P2 to P3.

While repeating such operations, the calculation parameters are tuned to optimal values, and the process amount pv gradually converges to a constant value.

In the above embodiment, P is determined based on the waveform observation method and waveform analysis results in the parameter tuning means 3.
The method for calculating I calculation parameters is disclosed in the patent application 1986-53.
Although it is assumed that the method proposed in No. 494 will be implemented, it is not necessary to use these methods.
1-245203 may be used.

(Effects of the Invention) As described above in detail, according to the present invention, the waveform observation of the response signal from the process reflects the actually set calculation parameters, and it is possible to perform appropriate waveform a measurement. It is possible to always tune the optimum calculation parameters.

[Brief explanation of the drawing]

Figure 1 is a block diagram showing the basic configuration of the present invention, Figure 2 is a block diagram showing the basic configuration of the present invention.
The figure is a configuration block diagram showing one embodiment of the present invention, FIG. 3 is a flowchart showing an example of operation, and FIG. 4 is a P calculation parameter (proportional gain) for changes in process amount P
FIG. 5 is a diagram showing how the parameters are sequentially changed by the parameter tuning means, as disclosed in US Pat. No. 46'0232.
6 is a configuration block diagram of a self-tuning type adjustment device described in Publication No. 6 (Japanese Unexamined Patent Publication No. 61-245203). 1... Controlled object (process) 2... PI control means 3... Parameter tuning means 4... Monitoring means 5... Waveform observation instruction means Fig. 4, Te 5

Claims (1)

[Scope of Claims] PI control means for performing at least proportional (P) and integral (I) calculations on the deviation signal between the process amount from the controlled object and the control target value, and outputting the obtained manipulated variable to the controlled object; , a parameter tuning means for observing a response signal waveform of a process quantity or a deviation signal, and tuning a PI calculation parameter of the PI control means so that the observed waveform becomes a desired response target, monitoring means for monitoring whether the absolute value of the deviation signal falls within a predetermined value range or whether this absolute value increases beyond a predetermined value; and detecting a parameter change by the parameter tuning means, and detecting a parameter change. 1. A control device comprising: waveform observation instruction means for instructing waveform observation of a response signal waveform after a predetermined period of time has elapsed.