JPH1069301A - Pid regulator - Google Patents

Abstract

PROBLEM TO BE SOLVED: To automatically set a preferable PID constant from control response with high probability without using a tuning signal. SOLUTION: A subtracter 1 outputs the deviation of a measured value from a target value. A PID control arithmetic part 3 outputs a manipulated variable, obtained from the deviation through PID operation, to a controlled system 5. An actuation judgement part 7 instructs the actuation of step response when the deviation exceeds a limit value. A step response tuning part 9 calculates a 1st PID constant by a step response method. A tuning observation part 11 outputs the degree of assurance of the 1st PID constant. A response evaluation tuning part 13 outputs a 2nd PID constant by an expert method. A PID constant variation judgement part 15 evaluates the 1st and 2nd PID constants and the degree of assurance unless a PID constant which is already set is excellent, and determines a 3rd PID constant according to them to set the change to the PID control arithmetic part 3.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a PID controller, and more particularly, to a PID controller which obtains a PID constant from a control response (measured value) of a process without adding a special tuning signal to a control object and automatically changes and sets the PID constant. It relates to possible PID controller improvements.

[0002]

2. Description of the Related Art Conventionally, as a PID controller of this kind,
The so-called expert method of observing the measured value from the control object when the disturbance or the target value is changed, and modifying the PID constant so that a good response can be obtained based on the observation result. It is well known that a so-called step response method for automatically calculating and changing and setting a PID constant from a step response at the start of operation is used.

[0003]

However, in the above-described PID controller, in order to reach the optimum PID constant by the expert method, it takes not only the skill of the operator but also a certain amount of time until the optimum PID constant is obtained. Was needed. On the other hand, in order to obtain a PID constant with a certain probability by the step response method, it is necessary to add an ideal step-like signal to the manipulated variable and to add it to the control target. There is also a difficulty that an error easily occurs when the control target is not stable.

SUMMARY OF THE INVENTION The present invention has been made to solve such a conventional disadvantage, and does not apply a special tuning signal to an object to be controlled, for example, an ideal step-like signal. It is an object of the present invention to provide a PID controller that can automatically calculate and set a constant with a high probability.

[0005]

In order to solve such a problem, a PID controller according to the present invention includes a PID controller for changing a target value such that a deviation between a target value and a measured value from a control target exceeds a predetermined limit value. And a start determination unit for instructing the start of the step response method when the target value is changed, and a first PID based on the step response method from the measured value when the target value is changed by the start instruction. A step response tuning unit for calculating a constant, a tuning observation unit for outputting the degree of certainty of the first PID constant from the measurement results and measurement conditions used for the calculation, a PID constant used for the PID calculation, A response evaluation tuning unit that evaluates the target value and the measured value based on the expert method standard, and outputs a second PID constant obtained by correcting the PID according to the degree of evaluation; When evaluation of the PID constant is not good, their first PID constants, degree of certainty, the P
Based on the evaluation result of the ID constant and the second PID constant,
A PID constant change determination unit that determines a third PID constant for changing the PID constant.

In the present invention, when the first PID constant is not calculated, the second PID constant is changed to the third PID constant.
It is preferable to form the PID constant change judging section so as to output as an ID constant. Further, the present invention compares the first and second PID constants, and when the two correction directions are the same, adds the third PID constant to the first and second PID constants by adding the degree of certainty. Determining a constant, and when the correction direction is different, from the second PID constant to the third PID
It is also possible to form the PID constant change determining section so as to determine the constant.

[0007]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of a PID controller according to the present invention. In FIG.
The subtracter 1 outputs a deviation E between a target value SV and a measured value PV described later, and is connected to the PID control calculation unit 3. The PID control calculation unit 3 performs a PID calculation of the deviation E based on the set PID constant to output the manipulated variable MV, and is connected to the control target 5.

The subtracter 1 and the PID control calculator 3
Is a basic configuration of a conventional PID controller, and PID constants set in advance in the PID control calculation unit 3 are set at the start of operation or before a target value is changed, and can be arbitrarily changed and set. It has become. The control target 5 includes a measuring unit (not shown) that measures a change in temperature, humidity, pressure, flow rate, or the like in a known process in which temperature control, humidity control, pressure control, flow rate control, or the like is performed by the operation amount MV. Have,
It outputs the measured value PV measured by this measuring means, and is connected to the subtracter 1, the activation judging unit 7, and the like.

At the start of operation or when the target value is changed, the start judging section 7 compares the deviation E between the target value SV and the measured value PV with, for example, a proportional band (not shown) and performs the deviation E only for executing tuning. It determines whether or not there is a tuning, and outputs a tuning start instruction signal when tuning can be executed. The step response tuning unit 9, the tuning observation unit 11, the response evaluation tuning unit 13, and the PI
It is connected to the D constant change determination unit 15. In addition, when the deviation E can be tuned by the step response, the activation determination unit 7 can output a step response activation instruction signal for instructing the calculation of the first PID constant by the step response to the step response tuning unit 9. Have been.

The step response tuning section 9 measures the measured value when the first target value SV1 is changed from the first target value SV1 to the second target value SV2 as shown in FIG. The equivalent waste time L of the control target 5, the inclination θ of the maximum speed slope R, and the time TR from the change of the target value SV to the determination of the maximum speed slope _R are measured based on the PV, and based on these measurements, The first P is calculated by a known calculation method such as the Ziegler-Nichols method.
It has a function of calculating an ID constant [(Kp, Ti, Td) _S ], and is connected to the tuning observation unit 11 and the PID constant change determination unit 15.

The tuning observing section 11 receives an operation amount M based on a tuning start instruction signal as shown in FIG.
Along with this by observing V determines whether exceeds a predetermined limit value, the target value manipulated variable MV after is performed change SV measures the time T _MV until out of the limit value, the time T _MV And the equivalent PI from the step response tuning unit 9, the time T _R , and the maximum speed slope R, as shown in FIG.
It has a function of calculating how good the calculation result of the D constant is, and is connected to the PID constant change determination unit 15. That is, in FIG. 3, when the ratio of the value obtained by adding the equivalent waste time L to the time T _MV to the time T _R is larger than “1” in FIG. 3, the first PID constant is calculated by the step response. Is judged to be successful, and when the ratio is smaller than “1”, the first PI becomes smaller as the ratio becomes smaller.
It has a function of judging that the calculation of the D constant has not been gradually improved, and outputting a coefficient CF corresponding to the calculation to the PID constant change judgment unit 15.

The tuning observation unit 11 sets a plurality of speed measurement widths (..., Bn, Bn + 1,...) As shown in FIG. For example, the speed measurement width Bn at which the maximum inclination point can be obtained
The number of passing samples of the measured value PV in the above is counted and averaged, and a value corresponding to the number of samples passing through the speed measurement width Bn is taken as the degree of certainty of the maximum speed slope point R as shown in FIG. The unit 9 has a function of outputting to the PID constant change determination unit 15 together with the degree of certainty of the measurement result. For this reason, if the number of passing samples of the speed measurement width is large, the degree of certainty for the speed measurement result is high, and if the number of passing samples is small, the degree of certainty is low.
There is an advantage that measurement accuracy can be improved with respect to counter variations and noise during A / D conversion of V.

Returning to FIG. 1, the response evaluation tuning unit 1
3 is the target value S based on the tuning start instruction signal.
The response is evaluated by measuring the response speed, overshoot, and damping ratio of the controlled object 5 from V and the measured value PV. If the response is poor, the expert method is used based on the response evaluation result based on the response evaluation result. 2 has a function of calculating a PID constant [(Kp, Ti, Td) _E ] and outputting a response evaluation result and a second PID constant.

That is, the response evaluation tuning unit 13
Represents the response speed of the control target 5 as shown in FIG.
The measured value PV is changed from the first target value SV1 to the second target value SV.
2, a response corresponding to the value of Tb / Ta, where Tb is the time required to reach 2 and Ta is the time required for the measured value PV to reach ΔSV / 2 (ΔSV = SV2−SV1) from the first target value SV1. Evaluate using the speed evaluation coefficient. This response speed evaluation coefficient is “1” or more, indicating that the response speed is good. When the value of Tb / Ta exceeds 2 and becomes large, the evaluation coefficient becomes smaller than “1” and the evaluation decreases. If the evaluation is reduced, the evaluation coefficient is set to K
Let p be the correction factor.

As shown in FIG. 6B, the response evaluation tuning unit 13 calculates the difference between the first and second target values SV1 and SV2 for the overshoot of the control target 5 by ΔS
When V is V and the maximum overshoot value is OV1, it has a function of evaluating using an overshoot evaluation coefficient corresponding to the value of OV1 / ΔSV. This overshoot evaluation coefficient is “1” or more, indicating that the evaluation is good. When the value of OV1 / ΔSV exceeds 0.15 and becomes large, the evaluation coefficient becomes smaller than “1” and the evaluation decreases. When the evaluation is lowered, the evaluation coefficient is used as a correction coefficient for Ti and Td.

Further, the response evaluation tuning unit 13
As shown in FIG. 6C, when the first overshoot value is a1 and the second overshoot value is a2 as shown in FIG. 6C, the damping ratio of the controlled object 5 is determined by using an evaluation coefficient corresponding to the ratio between a2 and a1. Has the function of evaluating The attenuation evaluation coefficient is "1" or more, indicating that the attenuation ratio is good. When the value of a2 / a1 exceeds about 0.25 and becomes large, the evaluation coefficient becomes smaller than "1" and the evaluation decreases. When the evaluation decreases, the attenuation evaluation coefficient is used as a correction coefficient for Kp.

The response evaluation tuning unit 13
The correction coefficients corresponding to the response speed, overshoot, and damping ratio are evaluated. If all the evaluation coefficients are “1” or more, the PID constant set in the PID control calculation unit 3 is good. And if at least one of the evaluation coefficients is less than “1”, a second PID constant [(Kp) obtained by modifying the PID constants (Kp, Ti, Td) set in the PID control calculation unit 3 , Ti, Td) _E ] using the respective evaluation coefficients, and outputting the calculated coefficients to the PID constant change determination unit 15. In the present invention, as shown in FIG. 6, the reference of the response speed evaluation coefficient according to the value of Tb / Ta, the reference of the overshoot evaluation coefficient according to the value of OV1 / ΔSV, and the value of a2 / a1 A method of correcting and determining the PID constant using the reference of the corresponding attenuation evaluation coefficient is an expert method.

The PID constant change judging unit 15 operates based on the input of a tuning start instruction signal. When an evaluation result indicating that the PID constant is good is input from the response evaluation tuning unit 13, the PID control calculation is performed. The PID constant set in the unit 3 is not changed, and if an evaluation result indicating that the PID constant is not good is input, the first PID constant [(Kp, Ti, Td)] from the step response tuning unit 9 is input.
It is determined whether or not [ _S ] has been input. If the first PID constant has not been input, the second PID constant [(Kp, Ti, Td) _E ] corrected by the response evaluation tuning unit 13 is set to P.
The data is output to the ID control calculation unit 3 and the PID constants (Kp, Ti,
It has a function to replace Td).

Further, when the first PID constant [(Kp, Ti, Td) _S ] is input from the step response tuning unit 9, the PID constant change determination unit 15
The correction directions of the ID constant [(Kp, Ti, Td) _S ] and the second PID constant [(Kp, Ti, Td) _E ] are compared,
When the correction directions are different from each other, the second PID constant [(Kp, Ti, Td) _E ] is prioritized and is changed to the third PID constant.
The ID is output to the PID control calculation unit 3 as an ID constant, and when the correction directions are the same, the first PID constant [(Kp, T
i, Td) _S ] and a second PID constant [(Kp, Ti, T
to the difference between the d) _E], the second PID constant value corrected with certainty of certainty factor CF and speed maximum slope point R from the tuning observation section 11 [(Kp, Ti, Td ) E] And outputs the third PID constant to the PID control calculation unit 3. Correction by the certainty of the maximum speed slope point R is not essential.

In other words, the PID constant change determining unit 15
When the PID constant used in the PID control calculation unit 3 is not good, and when the degree of certainty of the first PID constant is high, the step response method is used in which the optimum value of the PID constant is obtained by changing the target value once. The first PID constant is defined as a third PID constant, and if the degree of certainty of the first PID constant is not high, the difference between the first PID constant and the second PID constant is determined by the difference. When a value obtained by adding the second PID constant to the value corrected by the degree coefficient CF is used as a third PID constant. Further, when tuning by the step response method cannot be executed and the first PID constant cannot be obtained, The second PID constant is determined as a third PID constant and its PI
It has a function to change the D constant.

Incidentally, the main part of the PID controller according to the present invention described above stores a CPU having the above-described calculation, determination and calculation processing functions, a ROM storing an operation program of the CPU, and stores calculation and calculation data. In many cases, the microcomputer is configured by a microcomputer having a RAM, an I / O as an interface, and the like. For example, the operation of the response evaluation tuning unit 13 described above is shown in a flowchart as shown in FIG. 7, and the operation of the PID constant change determination unit 15 is shown in FIG.
It is shown by a flowchart as shown in FIG.

That is, as shown in FIG. 7, when the program is started, the response evaluation tuning unit 13 sets "1" as a correction coefficient of Kp in step 100, and sets correction coefficients of Ti and Td in step 101. Is set to “1”, and it is determined in step 102 whether or not the response speed is good. If NO in step 102, the reciprocal of the response speed evaluation coefficient is set as a correction coefficient for Kp in step 103, and the process proceeds to step 104.
Is YES, the process directly proceeds to step 104, where it is determined whether or not the overshoot is good.

If step 104 is NO, the reciprocal of the overshoot evaluation coefficient is set to Ti, Td in step 105.
The process proceeds to step 106, and if step 104 is YES, the process directly proceeds to step 106 to determine whether the attenuation evaluation is good. Step 106
Is NO, the attenuation evaluation coefficient is set to Kp in step 107.
The correction coefficient is set as a correction coefficient, and the process proceeds to step 108. If step 106 is YES, the process directly proceeds to step 108. In step 108, the set value of Kp (previous set value) is multiplied by the correction coefficient of Kp to correct Kp. , T
The set values of i and Td (previous set values) are multiplied by the correction coefficients of Ti and Td to correct Ti and Td, and the process ends.

As shown in FIG. 8, when the program is started, the PID constant change judging section 15 judges the response evaluation of the target value, that is, whether the already set PID constant is good or not in step 200. Step 200 is YE
If S, the previously set PID constants (Kp, Ti,
End without replacing Td), step 20
If 0 is NO, the tuning in the step response tuning unit 9 in step 201, that is, the first PI
It is determined whether the calculation of the D constant has been performed. Step 2
If 01 is NO, the second PID constant [(K
p, Ti, Td) _E ] as the third PID constant
Output to the control calculation unit 3 and set the PID constant (K
(p, Ti, Td).

If the tuning in the step response tuning section 9 is executed and the step 201 is YES, the first PID constant [(Kp, T
i, Td) _S ] and a second PID constant [(Kp, Ti, T
d) It is determined whether the correction directions with _E ] are the same. If the correction directions are different from each other and the result is NO, in step 204, the second PID constant [(Kp, Ti, Td) _E ] is changed to the third. PID
The value is output to the PID control calculation unit 3 as a constant, and the processing ends. If the correction directions are the same and step 203 is YES, the first PID constant [(Kp, Ti, Td) _S ] and the second PID constant
The difference from the ID constant [(Kp, Ti, Td) _E ] is corrected by a certainty factor CF to a second PID constant [(K
p, Ti, Td) _E ] is added to the third PID constant
The output is output to the D control operation unit 3 and the processing is terminated.

As described above, the PID controller of the present invention detects whether the deviation E between the target value SV and the measured value PV is a target value change that exceeds a predetermined limit value by the start-up judging unit 7 and When it is determined that the change is the target value SV, a step response method start is instructed, and the start instruction causes the step response tuning unit 9 to change the first PID constant from the measured value PV at the time of the target value SV change by the step response method. The tuning observation unit 11 calculates the first PID
The degree of certainty of the constant is output, and the response evaluation tuning unit 13 uses the PID constant and the target value SV used in the PID calculation.
And the measured value PV is evaluated based on the expert method standard, and the PID constant is corrected according to the degree of the evaluation result to obtain the second P
An ID constant is output, and the PID constant change determination unit 15 outputs the first PID constant, the degree of certainty of the first PID constant,
Based on the evaluation result of the PID constant and the second PID constant, if the PID constant used for the PID calculation is not good, and if the degree of certainty of the first PID constant is high,
The first PID constant obtained by the step response method in which the optimum value of the PID constant is obtained by changing the target value once is converted to the third PI
When the degree of certainty of the first PID constant is not high, the difference between the first PID constant and the second PID constant is corrected to a value corrected by the certainty degree coefficient CF.
The value obtained by adding the constant is used as a third PID constant.
When the D constant is not obtained, the second PID constant is changed to the third PID constant.
Is determined and the PID constant is changed, and the target value SV and the measured value PV are evaluated without giving an ideal step-shaped tuning signal to the control target 5 in particular. The third PID constant, which is preferable only when it is not good, is automatically calculated with a high probability and can be changed as the PID constant used in the PID control calculation unit 3, so that the reliability of the PID constant change is improved.

Also, the first PID by the step response method
When the deviation E is such that the constant cannot be obtained, the third PID constant is determined based on the second PID constant by the expert method and the PID constant of the PID control calculation unit 3 is changed. Even if the first PID constant cannot be obtained, the security of the third PID constant to be changed can be secured. Further, the second PID constant by the expert method and the first PID constant by the step response method are evaluated, and the third PID constant is determined by preferentially using the second PID constant. Even if the measurement conditions of the constants are not perfect, it is possible to correct the second PID constant with a value obtained by adding the degree of certainty to the first PID constant, and to quickly obtain a more appropriate third PID constant. It becomes possible.

[0028]

As described above, the PID controller of the present invention can obtain a preferable PID constant from a process control response without giving a special tuning signal, for example, an ideal step-like signal, to a control object. Can be changed and set automatically. In addition, since the response is evaluated and the PID constant set in the control operation unit is evaluated, it is possible to change and set the PID constant only when the response is not good. Zu P
There is no problem that the ID constant is changed. When the first PID constant is not calculated, the PID constant change determination unit is configured to output the second PID constant as a third PID constant. Even if not obtained, the PID constant can be changed with the third PID constant having high security. Further, the first and second PID constants are compared, and when both correction directions are the same, the third PID constant is added to the first and second PID constants by adding the degree of certainty.
Determine a constant, and when the correction direction is different, its second PID
In the configuration in which the PID constant change determination unit is formed so as to output the constant as the third PID constant, a preferred third PID constant is formed according to the tendency of the first PID constant to be modified. Even though the tendency to correct the constant is preferred but not reliable, the degree of certainty and the second PI
A preferred third PID constant can be calculated quickly by taking into account the D constant.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of a PID controller according to the present invention.

FIG. 2 is a diagram showing the operation of the PID controller of FIG.

FIG. 3 is a diagram showing an operation of the PID controller of FIG. 1;

FIG. 4 is a diagram showing an operation of the PID controller of FIG. 1;

FIG. 5 is a view showing the operation of the PID controller of FIG. 1;

FIG. 6 is a diagram illustrating the operation of the PID controller of FIG. 1;

FIG. 7 is a flowchart showing the operation of the PID controller of FIG. 1;

FIG. 8 is a flowchart showing the operation of the PID controller of FIG.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Subtractor 3 PID control calculation part 5 Control object 7 Start-up judgment part 9 Step response tuning part 11 Tuning observation part 13 Response evaluation tuning part 15 PID constant change judgment part

Continued on the front page (72) Inventor Masashi Shimada 5-16-6 Kugahara, Ota-ku, Tokyo Inside Rika Kogyo Co., Ltd.

Claims (3)

[Claims] 1. A PID controller for calculating a PID based on a deviation between a target value and a measured value from a control target and outputting an operation amount to the control target, wherein the target value is such that the deviation exceeds a predetermined limit value. A start determination unit that detects whether or not the change is the target value and instructs a step response method start when the change is the target value. A step response tuning unit that calculates a PID constant of 1; a tuning observation unit that outputs the degree of certainty of the first PID constant from measurement results and measurement conditions used in the calculation; The PID constant, the target value, and the measured value are evaluated based on the expert method standard, and the PID constant used in the PID calculation is modified according to the degree of the evaluation result to obtain a second value. When the Response Evaluation tuning unit for outputting a PID constants, the evaluation of the PID constants used in the PID calculation is not good, the first from the step response tuning unit
The PID constant used in the PID calculation from the response evaluation tuning unit is changed based on the PID constant, the degree of the certainty from the tuning observation unit, the evaluation from the response evaluation tuning unit, and the second PID constant. A PID constant change determining unit that determines a third PID constant to be performed. 2. The method according to claim 1, wherein the PID constant change determination unit is configured to:
The PID controller according to claim 1, wherein the second PID constant is output as the third PID constant when the PID constant is not calculated. 3. The PID constant change judging section, wherein:
And the second PID constant is compared, and when both correction directions are the same, the third PID constant is determined by adding the degree of certainty to the first and second PID constants. 3. The PID controller according to claim 2, wherein when different, the third PID constant is determined from the second PID constant.