JP3312142B2 - Controller - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a controller such as a PID controller, and more particularly, to a controller for identifying a characteristic of a control target and setting a control parameter based on the characteristic.

[0002]

2. Description of the Related Art Conventionally, for example, as a method for setting PID control parameters of a PID controller, a step-like operation amount is added to a control target, a characteristic of the control target is identified from a response waveform, and PID control is performed based on the characteristic. There is a step response method for setting parameters.

[0003]

In this step response method, if the characteristics of the control target are not correctly identified for some reason, the PID control parameter is set to an incorrect value, and the control performance is reduced. Will be worse.

For example, as shown in FIG. 7, when the power supply of the heater to be controlled is delayed even though the power supply of the PID controller is turned on and the operation is started, such a case is considered. The dead time including the delayed period is identified, and the identification is made longer than the correct dead time, so that the PID control parameter is not set correctly.

[0005] Alternatively, as shown in FIG.
In the case where a momentary power failure occurs and the PID controller is reset after the power of the controller and the control target is turned on and the operation is started, the dead time until the momentary power failure occurs is ignored. And the identification is made shorter than the correct dead time, and the PID control parameter is not set correctly.

SUMMARY OF THE INVENTION The present invention has been made in view of the above points, and when the characteristics of a control target are not correctly identified by the step response method, erroneous control parameters are set and control is performed. The purpose is to prevent this from happening.

[0007]

In order to achieve the above-mentioned object, the present invention is configured as follows.

That is, the present invention according to claim 1 is a controller for identifying a characteristic of a controlled object by a step response method, a step manipulated variable output means for outputting a step manipulated variable, and application of the step manipulated variable. A step response identification unit that identifies a dead time and a maximum slope based on a change in the control amount of the control object with respect to the identification result determination unit that determines whether identification by the step response identification unit has been correctly performed ; Control amount obtained from the control target
PID that outputs the manipulated variable based on the deviation between the target and the target value
Control means and the switch for identifying characteristics of a control object.
Step operation amount from step operation amount output means is controlled
And can be obtained in response to changes in the control amount.
When the dead time and the slope reach a predetermined condition,
Stop applying the step operation amount and return to the PID control means.
Control system switching means for shifting to PID control by
When the result determination means determines that the identification was performed correctly
In addition, the dead time and the dead time identified by the
Calculate PID control parameters based on the maximum slope
Setting means for setting the PID control means and
Output of limit cycle operation amount
Force means for applying the limit cycle manipulated variable.
Based on the change in the control amount of the control target, the dead time and
And a limit cycle identification means for identifying the maximum slope
The control system switching means is the same as the identification result determination means.
When it is determined that the setting was not performed correctly,
Stop the PID control by the ID control means
Limit cycle manipulated variable from cycle manipulated variable output means
Is applied to the control object, and a virtual eye different from the target value is applied.
The limit cycle is performed on the standard value.

According to a second aspect of the present invention, in the controller of the first aspect, the identification result judging means is provided in the controller.
Product of the dead time and the maximum slope identified by the
Is correctly identified when it is within the predetermined range.
Is determined to have been made.

[0010]

According to the first aspect of the present invention, since the identification result determining means for determining whether the dead time and the maximum slope have been correctly identified by the step response method is provided,
When the identification by the step response method is not performed correctly, it is possible to prevent the control from being performed by setting the wrong control parameter, and to reduce the dead time and the maximum
If it is determined that the slope has been correctly identified,
Based on these, PID control parameters are set and P
ID control will be performed, and will continue to the step response
Therefore, optimal PID control is performed. further,
It was determined that identification by the step response method was not performed correctly.
Specified, it is identified by the limit cycle method.
The dead time identified by this limit cycle method
Optimum PID control parameters based on
Data can be set.

According to the present invention, the identification result is obtained.
In the determination means, the dead time identified by the step response method is
When the product with the maximum slope is within the predetermined range,
Is determined to be correct, for example,
When comparing only with a predetermined value to make a decision
In this case, the determination can be made more accurately than in the case of.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 is a functional block diagram of an embodiment of the present invention. In this embodiment, description will be made by applying to a PID controller.

The PID controller 1 of this embodiment is a self-adjustment type controller that automatically adjusts PID control parameters according to a control target 2.

The PID controller 1 of this embodiment identifies a rising characteristic of the controlled object 2 by a step response method when a target value is changed, and a dead time and a slope sequentially obtained at the time of the rising reach predetermined conditions. When you do
When the process shifts to the ID control, the PID control parameter is calculated based on the dead time and the maximum slope identified so far, and when it is determined that the identification by the step response method has been correctly performed, the calculated PID control parameter is set. If it is determined that the identification has not been correctly performed, the identification is immediately performed by the limit cycle method, and the PID control parameter is calculated and set based on the dead time and the maximum slope identified by the limit cycle method.

Therefore, in the PID controller 1, as shown in FIG. 1, a step operation amount output means 3 for outputting a step operation amount of 100% operation amount, and a PID controller for performing PID control.
Based on an ID control means 4, a limit operation amount output means 8 for outputting a limit operation amount that causes a response waveform of the control target 2 to vibrate, and a control amount change by a step response method,
The dead time Ln and the slope Rn corresponding to the change are sequentially calculated at a predetermined sampling cycle described later, and the dead time Ln is calculated.
An identification means 5 for identifying the dead time L and the maximum inclination R by the limit cycle method when the identification by the step response method is not correctly performed, and a step response method by the identification means 5. Identification result determining means 9 which determines whether or not the identification by the controller has been correctly performed and gives the identification result to the control system switching means 6 and the setting means 7
And the control system switching means 6 for switching and applying the step operation amount, the PID operation amount or the limit cycle operation amount to the control target 2 as described later, and the dead time L and the maximum slope R identified by the identification means 5. Setting means 7 for calculating PID control parameters based on the calculated PID control parameters and changing the PID control parameters of the PID control means 4 to the calculated PID control parameters.

The identification result determining means 9 determines whether or not the identification of the dead time L and the maximum slope R by the identifying means 5 by the step response method has been correctly performed. The product (L * R) of the dead time L and the maximum slope R identified in the above is within a predetermined range (L * R
When in _min ~L * R _max) is identified and determined to have been performed correctly, when not within the range, the identification is to determine that no correctly.

That is, in a normal control object, the dead time L and the maximum slope R have an inversely proportional relationship. Therefore, if the product (L * R) is correctly identified, the product is constant. It can be expected to fall within the range. Therefore,
In this embodiment, the lower limit value L * _Rmin and the upper limit value L * _Rmax are determined in advance experimentally or empirically as values that cannot be a normal control target.

According to the determination of the identification result of the step response method by the identification result determination means 9, it is determined that, for example, the identification as shown in FIG. 7 or FIG. .

As another embodiment of the present invention, the identification result judging means 9 may judge, for example, whether or not the identification has been correctly performed as follows.

If the identified dead time L exceeds a large value _Lmax which cannot be a predetermined normal control target, it is determined that the operation has not been correctly performed.

The identified dead time L and maximum slope R
If the product (L * R) exceeds the upper limit L * _Rmax , it is determined that the operation has not been performed correctly.

If the identified dead time L is less than a predetermined small value L _min that cannot be a normal control target, it is determined that the operation has not been correctly performed.

The identified dead time L and maximum slope R
If the product (L * R) is less than the lower limit L * R _min , it is determined that the operation was not performed correctly.

Even if the application of the step manipulated variable is continued, when the maximum gradient Rn cannot be obtained any more and the maximum gradient cannot be obtained (when the identification is completed before the maximum gradient is obtained), the operation can be correctly performed. It is determined that it has not been performed.

Determination is made by arbitrarily combining the above.

The control system switching means 6 applies a step operation amount to the control object 2 in order to identify the characteristics of the control object 2 and also uses the dead time Ln sequentially obtained from the identification means 5.
And when the slope Rn reaches a predetermined condition, the application of the step operation amount is stopped and the PID
Transfer to control.

Further, the dead time L by the step response method
When the identification result determination means 9 determines that the identification of the maximum slope R has not been correctly performed as described above, the PID control is stopped, and the limit cycle operation amount from the limit cycle operation amount output means 8 is immediately changed. Applied to the control target 2 and the dead time L
When the maximum slope R is identified by the identification means 5, P
This is to shift to ID control. At this time, the PID control parameters based on the dead time L and the maximum slope R identified by the limit cycle method are calculated by the setting means 7 and the PID control parameters of the PID control means 4 are changed.

In this embodiment, the identification means 5 has a function as a step response identification means for identifying the dead time L and the maximum slope R by a step response method as described later. Dead time L and maximum slope R by the limit cycle method when it is determined that identification by the step response method has not been performed correctly.
Has also a function as a limit cycle identification means for identifying.

The control system switching means 6 changes the target value,
In addition, when the current control amount is outside the proportional band, the self-tuning for setting the PID control parameter by switching the operation amount applied to the control target 2 to the maximum step operation amount from the step operation amount output means 3 Move to This step operation amount is applied when the dead time Ln and the slope Rn sequentially obtained from the identification unit 5 reach predetermined conditions, and in this embodiment, when the following equation (1) is no longer satisfied: The application of the step operation amount is stopped, and the flow shifts to PID control by the PID control means 4.

Current control amount (y) + slope (Rn ′) × dead time (Ln ′) <target value (1) Here, Rn ′ is the largest slope among slopes Rn calculated so far. Ln 'indicates a dead time corresponding to the slope Rn'.

That is, in this embodiment, the application of the step operation amount is continued within a range not exceeding the target value, and when the above equation (1) is no longer satisfied, the process shifts to PID control.

The identification means 5 calculates the control amount based on a change in the control amount due to the step response of the control object to the step operation amount.
The dead time Ln and the slope Rn corresponding to the change are sequentially calculated and identified as shown in FIG. 2, and output to the control system switching means 6, setting means 7, and identification result determination means 9.

FIG. 2A shows the maximum slope R and the dead time L when the initial state is equilibrium, and FIG. 2B shows the maximum slope R and the dead time Ln when the initial state is non-equilibrium. I have. In this embodiment, the slope Rn is calculated as an average value of eight samplings at a sampling period of 0.5 sec, the detection point of the slope Rn is set as the center point of the eight samplings, and the dead time Ln is determined by a step operation. It is calculated as the time from when the amount is applied to when the slope Rn intersects as shown in FIG.

As described above, the identification means 5 is used for 0.5 seconds.
The slope Rn and the dead time Ln are sequentially calculated and identified for each c.

The maximum gradient R is the maximum gradient at which a larger gradient Rn cannot be obtained even when the application of the step operation amount is continued. In this embodiment, the maximum gradient is the maximum gradient. Although it may be identified as R, as will be described later, even if the maximum slope has not yet been reached, the largest slope has not yet reached the maximum in order to make the transition to more preferable control parameters faster. It may be identified as the slope R.

Therefore, in this embodiment, even if the application of the step operation amount is continued, even if the maximum slope at which a larger slope Rn cannot be obtained is not identified as the maximum slope R, the identified waste is determined. The product (L * R) of the time L and the maximum slope R is within a predetermined range (L * R _min ~
L * R _max ), it is determined that identification has been correctly performed.

The setting means 7 of this embodiment determines P based on the maximum slope R and the dead time L identified by the identification means 5.
An ID control parameter is calculated, and when the identification result determination means 9 determines that the maximum slope R and the dead time L have been correctly identified by the step response method, the calculated PID control parameter is converted to a PID. Control means 4
And the new PID control parameter
ID control is performed.

When the identification result determination means 9 determines that the maximum slope R and the dead time L have not been correctly identified by the step response method, the calculated PI
Do not set D control parameters. In this case, as described above, the PID control is stopped, the process shifts to the limit cycle method, and the maximum slope R and the dead time L are identified by the limit cycle method. When the maximum slope R and the dead time L are identified by the limit cycle method, a PID control parameter is calculated based on the maximum slope R and the dead time L and set in the PID control means 4.

The above step manipulated variable output means 3, PID
The control means 4, the control system switching means 6, the identification means 5, the setting means 7, the limit cycle manipulated variable output means 8 and the identification result determination means 9 are constituted by microcomputers.

Next, the above operation will be described in more detail with reference to the flowcharts of FIGS.

First, it is determined whether or not the target value has been changed (step n1). When the target value has been changed, it is determined whether or not the current control amount is outside the proportional band (step n2).
When it is out of the proportional band, the process shifts to self-tuning and a step operation amount of 100% is applied (step n).
3) Based on the change in the control amount due to the step response of the control target 2, the dead time Ln and the slope Rn corresponding to the change are sequentially calculated and identified (step n4).

It is determined whether or not the dead time Ln and the slope Rn obtained by this identification satisfy the above-mentioned equation (1) (step n5). If so, the step operation amount is applied (step n3). And the identification (step n4) is continued, and if not satisfied, the application of the step operation amount is stopped to shift to the PID control (step n6), and the maximum slope R (the application of the step operation amount Even if it is continued, it is determined whether or not a maximum gradient (a maximum gradient at which a larger gradient Rn cannot be obtained) is obtained (step n7).

When the maximum gradient R is obtained, the gradient and the corresponding dead time are identified as the maximum gradient R and the dead time L, and a PID control parameter is calculated based on them (step n8).

Next, whether the maximum slope R and the dead time L have been correctly identified by the step response method is determined by determining whether the product R * L is within a predetermined range as described above. (Step n9), and when it is determined that the operation has been correctly performed, the PID control parameter of the PID control means 4 is changed to the PID control parameter calculated in Step n8 (Step n10), and the process ends.

In step n9, the maximum gradient R
When it is determined that the identification of the dead time L has not been correctly performed, the PID control is stopped, the limit cycle operation amount is output to the control target 2, and the identification by the limit cycle method is started, as shown in FIG. (Step n
11) It is determined whether or not the identification of the maximum slope R and the dead time L by the limit cycle method has been completed (step n).
12) When the identification is completed, a PID control parameter is calculated based on the maximum slope R and the dead time L identified by the limit cycle method, and the calculated P
While changing to the ID control parameter, the process shifts to the PID control (step n13) and ends.

When the maximum slope R is not obtained in step n7, the maximum slope Rn and the dead time Ln obtained so far are identified as the maximum slope R and the dead time L, and A PID control parameter is calculated based on the PID control parameter (step n14).
It is determined whether the gain is greater than the proportional gain of the D control parameter (step n15).

In step n15, when it is determined that the proportional gain is not large, that is, it is on the stable side, the maximum slope R and the dead time L by the step response method are determined.
It is determined whether or not identification has been correctly performed (step n1).
6) If it is determined that the processing has been performed correctly, the PID control parameter of the PID control means 4 is changed to the PID control parameter calculated in step n14 (step n).
17) The process proceeds to step n18, and if it is determined in step n16 that the operation was not performed correctly, the above-described step n
Move to 11.

When it is determined in step n15 that the proportional gain is large, the process proceeds to step n18 without changing the PID control parameters as shown in FIG.

In step n18, after the application of the step operation amount, the slope Rn and the dead time Ln are calculated until the dead time Ln elapses, and the maximum slope Rn and the dead time Ln up to that time are calculated as the maximum slope. R and the dead time L, and based on them, a PID control parameter is calculated (step n19), and its proportional gain Kp
Is determined to be larger than the proportional gain of the PID control parameter set in the PID control means 4 (step n20). When it is determined that the proportional gain is not large, the maximum slope R and the waste It is determined whether or not the time L has been correctly identified (step n2).
1) If it is determined that the operation has been performed correctly, the PID control parameter of the PID control means 4 is changed to the PID control parameter calculated in step n19 (step n
22) Then, when it is determined that the operation has not been correctly performed, the process proceeds to step n11.

When it is determined in step n20 that the proportional gain is large, it is determined that the already set PID control parameter is more suitable, and the process ends without changing the PID control parameter.

As described above, the dead time L and the maximum slope R are identified from the rising characteristics by the step response method, and the process shifts to the PID control. In addition, it is determined whether or not the identification by the step response method is correctly performed. When it is determined that the control has been performed, the PID control parameter is changed based on the identified dead time L and the maximum slope R to shift to the optimal control. When it is determined that the control has not been performed correctly, the limit cycle method is immediately performed. The dead time L and the maximum slope R are identified again by the
The control is shifted to the optimal control by changing the PID control parameter on the basis of the PID control parameter. Therefore, even if the characteristic identification by the step response method fails as shown in FIG. 7 or FIG. The control can be shifted to the optimal control without deteriorating.

Further, according to this embodiment, since the process shifts to the PID control following the step response, it is not necessary to start a test for characteristic identification other than the normal closed loop control.

Further, in this embodiment, even if the application of the step operation amount is continued, if the maximum inclination R at which a larger inclination Rn cannot be obtained cannot be obtained, the PI
After shifting to the D control, it is determined whether or not to change the PID control parameter by comparing the proportional gain, so that the PID control parameter is not changed to the unstable side, and therefore, the control system becomes unstable. It is hard to become.

In the above-described embodiment, the proportional gain is compared to determine whether to change the PID control parameter. However, as another embodiment of the present invention, the comparison of the proportional gain may be omitted. .

In the above-described embodiment, after once shifting to the PID control, if it is determined that the identification by the step response method has not been correctly performed, the process immediately shifts to the limit cycle method as shown in FIG. As shown in
After the controlled variable reaches the target value, the limit cycle method
Instead of shifting to the control amount, as shown in FIG.
Without waiting for the target to reach the target value,
The limit cycle is performed for the desired value.
You.

[0057]

According to the present invention as described above, according to the present invention, when the constant that by the step response method is not correctly performed, erroneous control parameters is rather named control is set is performed, when identification by step response method is correct, since control parameters are set based on it, so that optimum PID control is performed subsequent to the step response. Further, when the identification by the step response method is not correctly performed, the identification is performed by the limit cycle method, so that it is possible to set a control parameter based on the identification result and shift to the optimal control.

[Brief description of the drawings]

FIG. 1 is a functional block diagram of an embodiment of the present invention.

FIG. 2 is a waveform chart for explaining calculation of a slope Rn and a dead time Ln.

FIG. 3 is a flowchart for explaining the operation.

FIG. 4 is a flowchart for explaining the operation.

FIG. 5 is a waveform chart for explaining a transition to a limit cycle method.

FIG. 6 is a waveform chart for comparison with the present invention.

FIG. 7 is a waveform chart for explaining a problem of the conventional example.

FIG. 8 is a waveform chart for explaining a problem of the conventional example.

[Explanation of symbols]

 2 Control target 3 Step manipulated variable output means 4 PID control means 5 Identification means 6 Control system switching means 7 Setting means 8 Limit cycle manipulated variable output means 9 Identification result judgment means

──────────────────────────────────────────────────の Continuing on the front page (51) Int.Cl. ⁷ identification code FI G05B 23/02 G05B 23/02 X (58) Field surveyed (Int.Cl. ⁷ , DB name) G05B 11/36-11 / 42 G05B 13/02 G05B 23/02

Claims (2)

(57) [Claims] 1. A controller for identifying characteristics of a controlled object by a step response method, comprising: a step manipulated variable output means for outputting a step manipulated variable; and a change in a controlled variable of the controlled subject with respect to application of the step manipulated variable. Based on the step response identification means to identify the dead time and the maximum slope, identification result determination means to determine whether the identification by the step response identification means was performed correctly, and a control amount obtained from the control object Based on the deviation from the target value
PID control means for outputting the manipulated variable, and the step manipulated variable for identifying the characteristic of the controlled object.
When the step operation amount from the output means is applied to the control target
In both cases, the dead time and the
When the tilt and inclination have reached the specified conditions,
The application of the amount is stopped and the PID control is performed by the PID control means.
Control system switching means for shifting to control, and the identification result determination means determines that identification has been correctly performed.
When the step response identification means
PID control parameters based on dead time and maximum slope
Setting means for calculating data and setting the PID control means
And limit cycle operation that outputs the limit cycle manipulated variable.
Production output means, and the control pair for the application of the limit cycle manipulated variable.
Dead time and maximum slope based on changes in elephant control
Limit cycle identification means for identifying the control system , wherein the control system switching means, the identification by the identification result determination means,
When it is determined that the operation was not performed correctly, the PID
The PID control by the control means is stopped and the limit size
Before the limit cycle manipulated variable from the
A virtual target value applied to the control object and different from the target value.
A controller that performs a limit cycle on the controller. 2. The method according to claim 1, wherein the identification result determination means is configured to perform the step response.
The product of the dead time identified by the answer identification means and the maximum slope is
Identification was correctly performed when it was within the predetermined range
2. The controller according to claim 1, wherein the controller determines the following.