JPH06289909A - IMC controller - Google Patents

Abstract

(57) [Summary] [Object] To provide an IMC controller capable of performing highly accurate and reliable control even when hysteresis, non-linearity, or time variation is present in a process to be controlled. [Structure] The internal model storage unit 6a stores the parameters of the internal model, which is a mathematical expression of the controlled process. The characteristics of the target value filter unit 2 and the target value / disturbance filter unit inside the manipulated variable calculation unit 4 are determined by the first time constant and the second time constant, respectively. Then, the filter time constant selection processing unit 11 selects and outputs the first time constant and the second time constant corresponding to the control condition from the filter time constant candidates stored in the filter time constant candidate storage unit 10. It
Therefore, the characteristics of the target value / disturbance filter unit inside the target value filter unit 2 and the manipulated variable calculation unit 4 are changed according to the control conditions.

Description

Detailed Description of the Invention

[0001]

The present invention relates to an IMC (Internal Model)
The present invention relates to a controller using a control algorithm having a (Control) structure, and particularly to a controller capable of changing a time constant of a filter according to a control condition.

[0002]

2. Description of the Related Art Conventionally, a PID has been used as a general-purpose controller.
The one using control is generally used. The PID controller uses an operation unit that performs a PID operation to calculate an operation amount (indoor condition) that is an output of the controller from a difference between a target value (e.g., if the controller is an indoor air conditioner corresponds to an indoor temperature set value) and a feedback amount. A feedback control system that calculates the temperature of hot air or cold air from the air conditioner, outputs this manipulated variable to the controlled process (indoor environment), and returns the control result (indoor temperature) as the feedback amount. is there. However, in the PID controller, the time from the output of the manipulated variable to the change in the controlled variable in the process to be controlled (for example, in the case of an indoor air conditioner, the time from the release of warm air to the rise in the indoor temperature).
However, there is a problem that it is difficult to deal with a large dead time.

Therefore, there has been proposed a controller using an IMC structure control algorithm for performing control by incorporating an internal model in which a process to be controlled is expressed by a mathematical expression. FIG. 7 is a block diagram of a control system using this IMC controller. 43 is a first subtraction processing unit for subtracting a feedback amount described later from the target value, 42 is a filter unit for preventing a change in the output of the first subtraction processing unit 43 from being suddenly transmitted, and 44 is a filter unit 42. An operation unit that calculates an operation amount based on the output of the control target, 46 is an internal model that approximates the control target process by a mathematical expression, and outputs a reference control amount corresponding to the control amount of the control target process, and 48 is a control amount A second subtraction processing unit 50 that subtracts the reference control amount from the internal model 46 and outputs a feedback amount is a control target process. Further, F, Gc, Gm, and Gp are the filter unit 42, the operation unit 44, the internal model 46,
The transfer function of the controlled process 50, r is a target value, u is an operation amount, d is a disturbance corresponding to an outdoor environment with respect to an indoor environment, y is a control amount, ym is a reference control amount, and e is a feedback amount.

Next, the operation of such an IMC controller will be described. First, the first subtraction processing unit 43 subtracts the feedback amount e from the target value r, and the result is output to the filter unit 42. Next, the operation amount u is calculated by the operation unit 44 from the output of the filter unit 42, and is output to the control target process 50 and the internal model 46 of the controller. Then, the second subtraction processing unit 48 subtracts the reference control amount ym from the internal model 46 that approximates the control target process 50 from the control amount y of the control target process 50, and the result is the feedback amount e. A feedback control system that is fed back to the first subtraction processing unit 43 is configured as.

Ideally, the internal model 46 of such an IMC controller is expressed mathematically so as to be exactly the same as the controlled process 50. Normally, the controlled process 50 is gained, dead time, and time. It is expressed mathematically as having a characteristic consisting of a parameter called a constant.
Further, it is ideal that the operating section 44 has the inverse characteristic (1 / Gm) of the transfer function of the internal model 46, but since the element of the dead time in the internal model 46 cannot be reciprocal, Ignore the dead time factor.

Therefore, the control amount y can be obtained by the following equation from the target value r and the disturbance d with such a configuration. y = F * Gp * Gc * r / {1 + F * Gc * (Gp-Gm)} + (1-F * Gm * Gc) * d / {1 + F * Gc * (Gp-Gm)} ... (1 Here, the transfer function Gm of the internal model 46 is equal to the transfer function Gp of the control target process 50, and the transfer function Gc of the operating unit 44 is the reciprocal of the transfer function of the internal model 46 (1 / Gm =
Assuming an ideal state equal to 1 / Gp), equation (1) becomes y = F × r + (1-F) × d (2)

Furthermore, under ideal conditions in which the target value r does not change suddenly, the filter unit 42 is unnecessary and F = 1 can be set, so that the control amount y becomes equal to the target value r (y =
r), it is possible to realize the control without any influence of the disturbance d. Further, focusing on the disturbance d in the control system of FIG. 7, even if there is a large dead time in the controlled object process 50 and the internal model 46, both exhibit the same characteristics with respect to the manipulated variable u, so the second subtraction It can be seen that the feedback amount e which is the output of the processing unit 48 is only the disturbance d, and the disturbance d can be suppressed.

In such an IMC controller, normally, one type of filter unit 42 is designed based on design conditions such as robust stability indicating stability when the error between the controlled process 50 and the internal model 46 becomes large. To be done.
However, in the control target process 50, for example, when the outdoor temperature is 30 ° C. and the indoor temperature is decreased from 30 ° C. to 25 ° C., and when the indoor temperature is increased from 20 ° C. to 25 ° C., the method of changing the indoor temperature is different. There may be hysteresis having different characteristics when falling and when rising. Further, when the outdoor temperature is 30 ° C. and the indoor temperature is reduced from 30 ° C. to 25 ° C. and when the indoor temperature is reduced from 25 ° C. to 20 ° C., there is a nonlinearity in which the gain and the time constant change depending on the control condition. In some cases, there is a time-varying property in which the gain and time constant change depending on the passage of time, such as when controlling the indoor temperature for different outdoor temperatures in the morning and night.

When the control target process 50 has hysteresis, non-linearity, or time-varying property as described above, the characteristics of the control target process 50 change due to changes in the control conditions. Effective model identification errors result in changes in the robust stability of this controller. In the above case,
There is a method of designing the filter unit 42 by assuming a virtual model identification error of the internal model 46. However, considering the model identification error based on the characteristic change of the control target process 50, the time constant of the filter unit 42 is extremely large. It will be a big one.

[0010]

Since the conventional IMC controller is configured as described above, when there is hysteresis, non-linearity, or time variation in the process to be controlled, the robust stability of this controller changes. When the robust stability becomes insufficient, the control becomes unstable. In addition, when designing by assuming the actual model identification error of the internal model as described above, the time constant of the filter section becomes extremely large, and there is a problem that the target value followability deteriorates and practical problems occur. there were. In order to solve the above problems, the present invention is capable of performing highly accurate and reliable control even if hysteresis, non-linearity, or time variation exists in a controlled process.
It is intended to provide a controller.

[0011]

According to the present invention, a target value filter section for outputting an input target value with a characteristic determined by a first time constant, and a feedback amount is subtracted from the output of the target value filter section. A first subtraction processing unit, a target value / disturbance filter unit that outputs the output of the first subtraction processing unit with a characteristic determined by a second time constant, and a target value / disturbance filter based on the parameters of the internal model. An operation amount calculation unit including an operation unit that calculates an operation amount from the output of the unit, an internal model storage unit that stores parameters of the internal model,
An internal model output calculation unit that calculates a reference control amount from the operation amount output from the operation amount calculation unit based on the parameters of the internal model, and a reference control amount output from the internal model output calculation unit from the control amount of the controlled process It is assumed that the second subtraction processing unit that subtracts the output of the control signal and outputs the feedback amount, the condition signal input unit that outputs the control condition at the time of operation as the control condition signal, and the change of the characteristic of the controlled process due to the change of the control condition. A filter time constant candidate storage unit that stores the first time constant and the second time constant as a filter time constant candidate, and a filter time constant candidate storage unit that stores the filter time constant candidate storage unit based on the control condition signal output from the condition signal input unit. The filter time constant selection processing unit that selects and outputs the first time constant and the second time constant according to the control condition from the selected candidates.

A target value filter unit for outputting the input target value with a characteristic determined by the first time constant,
The output of the first subtraction processing unit and the first subtraction processing unit that subtracts the feedback amount from the output of the target value filter unit
And a manipulated variable calculation unit consisting of a target value / disturbance filter unit that outputs with the characteristics determined by the time constant of , An internal model storage unit that stores parameters of the internal model, an internal model output calculation unit that calculates a reference control amount from an operation amount output from the operation amount calculation unit based on the parameters of the internal model, and a control of a control target process A second subtraction processing unit that subtracts the reference control amount output from the internal model output calculation unit from the amount and outputs a feedback amount; a condition signal input unit that outputs a control condition during operation as a control condition signal; A filter time constant representative storage unit that stores typical first time constants and second time constants assuming changes in the characteristics of the controlled process due to changes in conditions, Based on the control condition signal output from the signal input unit and the representative first time constant and second time constant stored in the filter time constant representative storage unit, the first time constant according to the control condition and And a filter time constant estimator that estimates and outputs a second time constant.

Further, instead of the filter time constant estimation unit,
A membership function in fuzzy inference is set based on the representative first time constant and second representative time constant stored in the filter time constant representative storage unit, and the membership function and the condition signal input unit output the membership function. A filter time constant estimation processing unit that estimates and outputs a first time constant and a second time constant according to the control condition based on the control condition signal.

A target value filter section for outputting the input target value with a characteristic determined by the first time constant,
The output of the first subtraction processing unit and the first subtraction processing unit that subtracts the feedback amount from the output of the target value filter unit
And a manipulated variable calculation unit consisting of a target value / disturbance filter unit that outputs with the characteristics determined by the time constant of , An internal model storage unit that stores parameters of the internal model, an internal model output calculation unit that calculates a reference control amount from an operation amount output from the operation amount calculation unit based on the parameters of the internal model, and a control of a control target process A second subtraction processing unit that subtracts the reference control amount output from the internal model output calculation unit from the amount and outputs a feedback amount; a condition signal input unit that outputs a control condition during operation as a control condition signal; Filter time constant calculation processing for calculating and outputting a first time constant and a second time constant according to the control condition based on the control condition signal output from the signal input unit And it has a door.

[0015]

According to the present invention, the target value is input to the target value filter unit whose characteristics are determined by the first time constant, and the feedback amount is subtracted from the output of the target value filter unit in the first subtraction processing unit. . The result of this subtraction is output to the target value / disturbance filter unit in the manipulated variable calculation unit whose characteristics are determined by the second time constant, and the manipulated variable is calculated from this result in the operation unit, and the controlled object process and internal model output calculation unit Is output to. Next, the second subtraction processing unit subtracts the reference control amount from the internal model output calculation unit from the control amount of the control target process, and the result is fed back to the first subtraction processing unit as a feedback amount. Is configured. Then, the filter time constant selection processing unit selects and outputs the first time constant and the second time constant corresponding to the control condition from among the filter time constant candidates stored in the filter time constant candidate storage unit. Due to
The characteristics of the target value filter unit and the target value / disturbance filter unit are changed.

Further, the first time constant and the second time constant corresponding to the control condition are based on the representative first time constant and second representative time constant stored in the filter time constant representative storage unit by the filter time constant estimating unit. A time constant of 2 is estimated and output.

Further, the filter time constant estimation processing unit sets the membership function in the fuzzy inference based on the representative first time constant and second representative time constant stored in the filter time constant representative storage unit, The first time constant and the second time constant according to the control conditions are estimated and output.

Further, the filter time constant calculation processing unit calculates and outputs the first time constant and the second time constant according to the control conditions.

[0019]

1 is a block diagram of an IMC controller showing an embodiment of the present invention, and FIG. 2 is a block diagram of a control system using this IMC controller. In FIG. 1, 1
Is a target value input section for inputting a target value r set by an operator (not shown) to this controller, and 2 is a first value output from a filter time constant selection processing section for the target value r from the target value input section 1. A target value filter unit 3 that outputs with a characteristic determined by a time constant is a first subtraction processing unit that subtracts the feedback amount e from the output of the target value filter unit 2, and 4 is an output from a filter time constant selection processing unit that will be described later. The operation amount calculation unit 5 calculates the operation amount u from the output of the first subtraction processing unit 3 based on the generated second time constant and the parameter output from the internal model storage unit described later, and 5 is the operation amount calculation unit. 4 is a signal output unit that outputs the manipulated variable u output from the control unit 4 to a control target process (not shown in FIG. 1).

Further, 6a is an internal model storage unit for storing the parameters of the internal model of this IMC controller, 6a
b is a reference control amount ym that is calculated as an internal model based on the parameters output from the internal model storage unit 6a.
, An internal model output operation unit for outputting the control amount y from the controlled object process to the IMC controller, and an internal model output operation for the control amount y output from the control amount input unit 7. The second subtraction processing unit subtracts the reference control amount ym output from the unit 6b and outputs the feedback amount e.

Further, 9 is a condition signal input unit for outputting a control condition signal based on the target value r output from the target value input unit 1 and the control amount y output from the control amount input unit 7, and 10 is a control condition. It is a filter time constant candidate storage unit that stores a filter time constant candidate assuming a change in the characteristics of the control target process 50 due to the change of. A filter time constant selection processing unit 11 selects a corresponding candidate from the candidates stored in the filter time constant candidate storage unit 10 based on the control condition signal output from the condition signal input unit 9, and First
The characteristics of the target value filter unit 2 and the target value / disturbance filter unit, which will be described later, in the manipulated variable calculation unit 4 are changed by outputting the time constant of and the second time constant.

In FIG. 2, reference numeral 4a is inside the manipulated variable calculating unit 4, and the output of the first subtraction processing unit 3 is determined by the second time constant output from the filter time constant selection processing unit 11. Target value / disturbance filter unit that outputs characteristics 4
b is also inside the target value / disturbance filter unit 4
An operation unit for calculating the manipulated variable u from the output of a, an internal model 6 including an internal model storage unit 6a and an internal model output operation unit 6b, F1 a transfer function of the target value filter unit 2, and F2.
Is a transfer function of the target value / disturbance filter unit 4a. 2 includes the original controller excluding the condition signal input unit 9, the filter time constant candidate storage unit 10, and the filter time constant selection processing unit 11 from the IMC controller of FIG. 1, the control target process 50, and the disturbance d. It was rewritten as a control system.

Next, the operation of such an IMC controller will be described. The target value r is set by the operator or the like of this IMC controller, and is input to the target value filter unit 2 via the target value input unit 1. The first time constant output from the filter time constant selection processing unit 11 described later is T
When set to 1, the target value filter unit 2 outputs the target value r with the characteristic of the transfer function F1 determined by the first time constant T1 as in the following equation. F1 = 1 / (1 + T1 × s) (3) Next, the first subtraction processing unit 3 uses the target value filter unit 2
The feedback amount e output from the second subtraction processing unit 8 is subtracted from the output of.

When the second time constant output from the filter time constant selection processing unit 11 is T2, the target value / disturbance filter unit 4a in the manipulated variable calculation unit 4 is provided in the first subtraction processing unit 3. The output is output with the characteristic of the transfer function F2 determined by the second time constant T2 as the following equation. F2 = 1 / (1 + T2 × s) (4)

Similarly, the internal operation unit 4b calculates the operation amount u from the output of the target value / disturbance filter unit 4a, and the transfer function Gc of the operation amount u of the internal model 6 output from the internal model storage unit 6a. The following formula is obtained by the gain and the time constant, and is the reciprocal of the transfer function Gm of the internal model 6 excluding the elements of the dead time Lm as in the example of FIG. 7. Gc = (1 + Tm × s) / Km (5) Here, Km and Tm are gains of the internal model 6, respectively.
It is a time constant.

Therefore, the transfer function of the operation amount computing section 4 as a whole is given by the following equation. F2 × Gc = (1 + Tm × s) / {Km × (1 + T2 × s)} (6) In this way, the manipulated variable u is calculated from the output of the first subtraction processing unit 3 and the signal output unit It is output to the control target process 50 via 5 and is also output to the internal model output calculation unit 6b.

Next, the internal model 6 uses the gain Km, the dead time Lm, and the time constant Tm stored in the internal model storage unit 6a as the parameters to control the process 50 to be controlled by the primary delay and the dead time. This is a mathematical expression expressed as having elements, and the reference control amount ym is calculated from the operation amount u output from the operation amount operation unit 4 in the internal model output operation unit 6b. The transfer function Gm is given by the following equation. Gm = Km × exp (−Lm × s) / (1 + Tm × s) (7)

Next, the second subtraction processing unit 8 subtracts the reference control amount ym from the internal model output operation unit 6b from the control amount y from the controlled object process 50 input via the control amount input unit 7. Then, the feedback amount e is output. Then, this feedback amount e is input to the first subtraction processing unit 3 as described above. This completes the feedback control system consisting of this IMC controller.

In such a control system, when the controlled object process 50 has a hysteresis having different characteristics when the controlled variable y falls and when it rises, the condition signal input unit 9, the filter time constant candidate storage unit 10, and the The filter time constant selection processing unit 11 performs the following operations to set the target value filter unit 2
Also, the characteristics of the target value / disturbance filter unit 4a can be changed.

First, the condition signal input unit 9 compares the target value r output from the target value input unit 1 with the control amount y output from the control amount input unit 7. Then, when the target value r is larger than the control amount y, it can be determined that the control amount y increases due to the control for making the control amount y reach the target value r thereafter, so that the change in the control amount y is positive. Then, a control condition signal indicating that is output. Further, when the target value r is smaller than the control amount y, it can be determined that the control amount y will decrease thereafter, so a control condition signal indicating that the change in the control amount y is negative is output.

Next, in the filter time constant candidate storage unit 10, for example, as a candidate of two kinds of filter time constants, a first time constant assuming a change in the characteristic of the controlled process 50 when the controlled variable y increases. The filter time constant of T1a is 10 seconds, the second time constant T2a is 10 seconds, and the first time constant T1b is 20 seconds and the second time constant T2b is 20 seconds when the control amount y is decreasing. The constant is stored. In this example, when the control amount y rises, the model identification error is small, and the filter time constant is made small by emphasizing the target value followability, and when the control amount y falls, the model identification error is large and robust stability is ensured. To do this, the filter time constant is increased.

Then, the filter time constant selection processing section 11
When the control condition signal indicating that the change in the control amount y is positive is output from the condition signal input unit 9, the control amount is determined as a candidate from the candidates stored in the filter time constant candidate storage unit 10. The filter time constant when y increases is selected, and the first time constant T1a and the second time constant T2a thereof are set to the target value filter unit 2 and the target value / disturbance filter unit 4, respectively.
output to a. Therefore, the characteristics of the target value filter unit 2 and the target value / disturbance filter unit 4a are determined.

Further, after the control amount y is increased by such control, for example, if the control amount y once overshoots and becomes larger than the target value r, it is shown that the change of the control amount y is negative. A control condition signal is output from the condition signal input unit 9, and the first time constant T1b and the second time constant T1b for decreasing the control amount y stored in the filter time constant selection processing unit 11 are stored in the filter time constant candidate storage unit 10. Is output, and the characteristics of the target value filter unit 2 and the target value / disturbance filter unit 4a are changed. Thereafter, such processing is always performed during the control.

Therefore, the IMC controller of this embodiment changes the characteristics of the target value filter unit 2 and the target value / disturbance filter unit 4a according to the control conditions, so that the filter time constant is not increased more than necessary. , Reliable control can be performed. Further, in the present embodiment, such a change of the filter time constant is applied when the control target process 50 has hysteresis, but it is also applied when the control target process 50 has nonlinearity or time variation. can do.

The control system based on the IMC controller of this embodiment is similar to the control system of the example of FIG.
2 corresponds to the control system in which the target value / disturbance filter unit 4a is set and the target value filter unit 2 is added to the target value r. Therefore, the control amount y is given by the following expression from Expression (1). y = F1 * F2 * Gp * Gc * r / {1 + F2 * Gc * (Gp-Gm)} + (1-F2 * Gm * Gc) * d / {1 + F2 * Gc * (Gp-Gm)} ... (8)

That is, as shown in equation (8), the second term on the right side of the disturbance d [(1-F2 × Gm × Gc) × d /
Since only the transfer function F2 of the target value / disturbance filter unit 4a is related to {1 + F2 × Gc × (Gp−Gm)}], the target value / disturbance filter unit 4a adjusts the disturbance d during design. To do. Also, the first term on the right side [F1 × F2 × G
p * Gc * r / {1 + F2 * Gc * (Gp-Gm)}]
Therefore, regarding the target value r, the target value / disturbance filter unit 4a
The target value filter unit 2 is adjusted after the adjustment of. That is, the IMC controller has a transfer function F2 having a first-order delay characteristic with respect to the disturbance d, and a transfer function F1 × F2 having a second-order delay characteristic with respect to the target value r.

In the example of FIG. 1, the pre-stored filter time constant candidates are selected according to the control conditions to change the characteristics of the filter section. However, from the typical filter time constants assuming the characteristics of the process to be controlled are changed. It is also possible to estimate and change the filter time constant according to the control condition. FIG. 3 is a block diagram of an IMC controller showing another embodiment of the present invention, and the same parts as those in FIG. 1 are designated by the same reference numerals.

Reference numeral 19 is a condition signal input unit for outputting a control condition signal based on the control amount y output from the control amount input unit 7, and 20 is a control target process 50 due to a change in the control condition.
It is a filter time constant representative storage unit that stores a typical filter time constant assuming a change in the characteristic of. Reference numeral 21 denotes a filter time constant estimation unit, which controls the control condition signal output from the condition signal input unit 19 and the filter time constant representative storage unit 20.
The target value filter unit 2 and the target value / disturbance filter unit 4 are estimated by outputting a filter time constant corresponding to the control condition based on the representative filter time constant stored in
Change the characteristics of a.

The IMC controller of this embodiment is the same as the example of FIG. 1 except for the condition signal input unit 19, the filter time constant representative storage unit 20, and the filter time constant estimation unit 21. Process 50
The operation of the condition signal input unit 19, the filter time constant representative storage unit 20, and the filter time constant estimation unit 21 when there is nonlinearity in which the gain and time constant change depending on the control condition will be described.

First, the condition signal input unit 19 sets the ratio of the control amount y output from the control amount input unit 7 to the preset full scale (hereinafter abbreviated as FS) of the control amount y as a control condition, This is output as a control condition signal. Next, the filter time constant representative storage unit 20 stores a representative filter time constant designed on the assumption that the characteristic of the controlled process 50 changes due to the change of the control condition.

Here, as typical filter time constants, for example, the control conditions are 0% FS, 50% FS, 100% FS.
Assuming that the time constant of the filter is stored, the control condition is 0%.
The first time constant T1 ₀ during FS is 10 seconds, and the second time constant T is
2 ₀ was 15 seconds, the first time constant T1 ₅₀ at 50% FS
Is 15 seconds and the second time constant T2 ₅₀ is 20 seconds, and 100%
The first time constant T1 ₁₀₀ during FS is 17 seconds and the second time constant T2 ₁₀₀ is 22 seconds.

By the way, since these are filter time constants for the above specific control conditions, for example, 0 to 50% FS.
Unstored filter time constants for control conditions between must be estimated. Such a filter time constant for a certain control condition sets a fuzzy set centered around the above specific control condition that indicates how close this filter time constant is to the above-mentioned representative filter time constant. It is estimated by fuzzy inference by setting the membership function that gives the goodness of fit, which is the degree to which belongs to the fuzzy set.

Then, the filter time constant estimation unit 21
For such a fuzzy inference, a specific control condition corresponding to a representative filter time constant stored in the filter time constant representative storage unit 20 is centered, that is, the adaptability 1.
A membership function that sets the degree of conformity of the control condition to 0 is set. FIG. 4 is an example of the membership function set in the filter time constant estimation unit 21, and here, three triangles having a matching degree of 1.0 with control conditions of 0% FS, 50% FS, and 100% FS, respectively. The membership function of is set.

Now, assuming that the control condition indicated by the control condition signal output from the condition signal input unit 19 is 30% FS at some point during control execution, the degree of conformity of the control condition 30% FS to each fuzzy set is shown in FIG. become that way. That is,
The goodness of fit α 0 for a fuzzy set centered on the control condition 0% FS (hereinafter abbreviated as a fuzzy set for the control condition 0% FS) is 0.4, and the goodness of fit α 50 for the fuzzy set with the control condition 50% FS is 0. The goodness of fit α100 (not shown) for a fuzzy set of 6,100% FS is zero.

Then, the filter time constant estimating unit 21 calculates the filter time constant when the control condition is 30% FS from the goodness of fit and the typical filter time constant stored in the filter time constant representative storage unit 20, that is, the first filter time constant. The time constant T1c of 1 and the second time constant T2c are estimated by the following equations. T1c = T1 ₀ × α0 + T1 ₅₀ × α50 + T1 ₁₀₀ × α100 = 10 × 0.4 + 15 × 0.6 + 17 × 0 = 13 seconds ... (9) T2c = T2 ₀ × α0 + T2 ₅₀ × α50 + T2 ₁₀₀ × α100 = 15 × 0.4 + 20 × 0.6 + 22 × 0 = 18 seconds (10)

Accordingly, the first time constant T1c and the second time constant T2c are output to the target value filter unit 2 and the target value / disturbance filter unit 4a, respectively, so that the target value filter unit 2 and the target value / disturbance are generated. The characteristics of the filter unit 4a are changed. Therefore, the IMC controller of this embodiment is
Since the characteristics of the target value filter unit 2 and the target value / disturbance filter unit 4a are changed according to the control conditions, it is possible to perform highly reliable control without increasing the filter time constant more than necessary. In the present embodiment, one type of control condition is used, but there may be two or more types of control conditions due to the characteristics of fuzzy reasoning, and the change of the filter time constant has non-linearity in the controlled process 50. Although it is applied to the case, it can also be applied to the case where the controlled object process 50 has hysteresis or time variation.

In the example of FIG. 3, the filter time constant is estimated by the filter time constant estimating unit 21 in which the membership function is set. However, in setting, adding or deleting such a membership function, a mathematical operation is performed. It may not be coordinated with the operator of the IMC controller as it may require knowledge. Therefore, a filter time constant estimation processing unit having a function of automatically setting a membership function from a typical filter time constant is provided instead of the filter time constant estimation unit 21 so that the membership function does not depend on the adjustment work. Can be set.

The configuration of the IMC controller in this embodiment is exactly the same as that of the example of FIG. 3 except for the filter time constant estimation processing unit, so that the operation will be described as the filter time constant estimation processing unit 21a. First, the filter time constant estimation processing unit 21a is provided with a general expression of a membership function such as the following expression. αA _n-1 = (C-A _n-1 ) / (A _n-1 -A _n ) +1 (11) αA _n = (C-A _n ) / (A _n -A _n-1 ) +1 ... (12) A _n-1 ≤ C ≤ A _n {n-1 = 1, 2. ． ． } ... (13)

Here, C is the control condition indicated by the control condition signal output from the condition signal input unit 19 of FIG. 3, A _n-1 , A _n
n is a control condition of a typical filter time constant stored in the filter time constant representative storage unit 20, and αA _n-1 and αA _n are goodness of fit to a fuzzy set centered around the control conditions A _n-1 and A _n , respectively. Is.

Now, assuming that the same representative filter time constants as in the example of FIG. 3 are stored in the filter time constant representative storage unit 20, the control conditions corresponding to the three types of filter time constants are A1 = 0, A2 = 50 and A3 = 100. Therefore, the control condition is 0 to 50% from the equations (11) and (12).
Goodness of fit α between FSs (0 ≦ C ≦ 50 in equation (13))
A1 and αA2 are as follows. αA1 = α0 = (C-A1) / (A1-A2) + 1 = -C / 50 + 1 (14) αA2 = α50 = (C-A2) / (A2-A1) + 1 = C / 50 ... (15)

Similarly, the control condition is 50-100% FS.
The degree of conformity α between (50 ≦ C ≦ 100 in equation (13))
A2 and αA3 are as follows. αA2 = α50 = (C−A2) / (A2−A3) + 1 = −C / 50 + 2 (16) αA3 = α100 = (C−A3) / (A3−A2) + 1 = C / 50-1 .. (17)

Equation (14) is the control condition 0% F in FIG.
This means a membership function of S, and the expression (15) is a control condition 50% FS membership function (control condition 50%
(Part below FS), formula (16) is the same 50% FS
The above part, Expression (17), means the membership function under the control condition of 100% FS. Now that the same membership function as in FIG. 4 has been set, the filter time constant can be estimated and changed in the same manner as in the example of FIG.

FIG. 5 shows the filter time constant estimation processing unit 2
It is another example of the membership function set by 1a, and in the example of FIG.
Since the estimation accuracy may be insufficient due to a valley between two membership functions, a filter time constant assuming the characteristic of the controlled process 50 under the control condition of 75% FS is additionally stored in the filter time constant representative storage unit 20. By
In this example, the membership function is reset as shown in FIG. Therefore, the IMC controller of the present embodiment can estimate the filter time constant simply by setting a typical filter time constant, so that the operator's burden of adjustment work can be reduced.

In the above embodiment, the filter time constant is estimated based on the control condition, but the filter time constant can be calculated from the control condition. FIG. 6 is a block diagram of an IMC controller showing another embodiment of the present invention.
The same parts as those in are denoted by the same reference numerals. Reference numeral 29 is a condition signal input unit that outputs the elapsed time from the start of control as a control condition signal, and 31 is a filter time constant based on the control condition signal output from the condition signal input unit 29. The target value filter unit 2
And a filter time constant calculation processing unit for changing the characteristics of the target value / disturbance filter unit 4a.

The IMC controller of this embodiment has the same configuration as that of the example of FIG. 1 except for the condition signal input section 29 and the filter time constant calculation processing section 31, so that the control target process 50 can be used with time. The operations of the condition signal input unit 29 and the filter time constant calculation processing unit 31 when there is a time variation in which the gain and the time constant change depending on each other will be described.

First, the condition signal input unit 29 sets the elapsed time t from the start of control as a control condition and outputs this as a control condition signal. Then, the filter time constant calculation processing unit 31
Calculates the first time constant T1d and the second time constant T2d at the elapsed time t indicated by the control condition signal output from the condition signal input unit 29, for example, as in the following equation. T1d = 5 × sin (π × t / 30) +10 (18) T2d = 7 × sin (π × t / 30) +15 (19)

In this example, the first time constant T1d and the second time constant T2d increase or decrease in a 60-minute cycle in order to ensure robust stability against changes in the characteristics of the controlled process 50. ing. At this time, assuming that the elapsed time t indicated by the control condition signal output from the condition signal input unit 29 is 250 minutes, the first time constant T1d and the second time constant T2d at the elapsed time t = 250 minutes are as follows. Calculated as shown. T1d = 5 × sin (250 × π / 30) + 10 = 14.3 seconds (20) T2d = 7 × sin (250 × π / 30) + 15 = 21.1 seconds (21)

Then, the filter time constant calculation processing unit 31 outputs the first time constant T1d and the second time constant T2d to the target value filter unit 2 and the target value / disturbance filter unit 4a, respectively. Filter unit 2 and target value
The characteristics of the disturbance filter unit 4a are changed. Therefore,
The IMC controller of this embodiment can perform highly reliable control by changing the filter time constant as described above, and its processing time is longer than that of the embodiment in which the filter time constant is estimated as in the example of FIG. It is short, and the required memory capacity can be reduced.

[0059]

According to the present invention, when the characteristic of the controlled process changes depending on the control condition, the filter time constant is selected, estimated, or calculated to be changed, so that accuracy and reliability are improved. High control can be performed. In addition, the filter time constant is not increased unnecessarily to deal with the model identification error in practice, and the control performance is not sacrificed at the same time as increasing the stability, so the IMC controller can be used for a wider range of applications. can do. Moreover, the membership function in fuzzy inference is set only by storing a representative filter time constant in this controller.
Since the filter time constant corresponding to the control condition is estimated, the operator's burden of adjustment work on this controller can be reduced. Also, since the filter time constant is calculated and changed according to the control condition, the processing time is short,
It is possible to reduce the required memory capacity. Further, in most practical cases, a high-grade processor or memory is not required, so that the cost can be reduced.

[Brief description of drawings]

FIG. 1 is a block diagram of an IMC controller showing an embodiment of the present invention.

FIG. 2 is a block diagram of a control system using the IMC controller of FIG.

FIG. 3 is a block diagram of an IMC controller showing another embodiment of the present invention.

FIG. 4 is an example of a membership function set in a filter time constant estimation unit in FIG.

FIG. 5 is an example of a membership function set by a filter time constant estimation processing unit.

FIG. 6 is a block diagram of an IMC controller showing another embodiment of the present invention.

FIG. 7 is a block diagram of a control system using a conventional IMC controller.

[Explanation of symbols]

 2 target value filter unit 3 first subtraction processing unit 4 manipulated variable calculation unit 4a target value / disturbance filter unit 4b operation unit 6a internal model storage unit 6b internal model output calculation unit 8 second subtraction processing unit 9 condition signal input unit 10 filter time constant candidate storage unit 11 filter time constant selection processing unit 19 condition signal input unit 20 filter time constant representative storage unit 21 filter time constant estimation unit 21a filter time constant estimation processing unit 29 condition signal input unit 31 filter time constant calculation processing Department

Claims (4)

[Claims] 1. A reference control amount corresponding to a control amount of a control target process, which is a control result, is calculated by calculating an operation amount to be output to a control target process from a control target value, and using an internal model expressing the control target process by a mathematical expression. Control is performed by calculating and feeding back the difference between the control amount and the reference control amount I
In the MC controller, a target value filter unit that outputs the input target value with a characteristic determined by a first time constant, a first subtraction processing unit that subtracts a feedback amount from the output of the target value filter unit, A target value / disturbance filter unit that outputs the output of the first subtraction processing unit with a characteristic determined by a second time constant, and an operation amount from the output of the target value / disturbance filter unit based on the parameters of the internal model. An operation amount calculation unit including an operation unit that calculates a value, an internal model storage unit that stores parameters of the internal model, and a reference control based on an operation amount output from the operation amount calculation unit based on the parameters of the internal model. The internal model output calculation unit that calculates the quantity and the reference control amount output from the internal model output calculation unit from the control quantity of the controlled object process A second subtraction processing unit that outputs a feedback amount, a condition signal input unit that outputs a control condition at the time of operation as a control condition signal, and a first subtraction processing unit that assumes a change in the characteristics of the control target process due to the change in the control condition. A filter time constant candidate storage unit that stores a time constant and a second time constant as a filter time constant candidate, and a filter time constant candidate storage unit that stores the filter time constant candidate storage unit based on a control condition signal output from the condition signal input unit. An IMC controller comprising: a filter time constant selection processing unit that selects and outputs the first time constant and the second time constant according to a control condition from the selected candidates. 2. A reference control amount corresponding to the control amount of the control target process, which is a control result, is calculated by calculating an operation amount to be output to the control target process from a control target value and using an internal model expressing the control target process by a mathematical expression. Control is performed by calculating and feeding back the difference between the control amount and the reference control amount I
In the MC controller, a target value filter unit that outputs the input target value with a characteristic determined by a first time constant, a first subtraction processing unit that subtracts a feedback amount from the output of the target value filter unit, A target value / disturbance filter unit that outputs the output of the first subtraction processing unit with a characteristic determined by a second time constant, and an operation amount from the output of the target value / disturbance filter unit based on the parameters of the internal model. An operation amount calculation unit including an operation unit that calculates a value, an internal model storage unit that stores parameters of the internal model, and a reference control based on an operation amount output from the operation amount calculation unit based on the parameters of the internal model. The internal model output calculation unit that calculates the quantity and the reference control amount output from the internal model output calculation unit from the control quantity of the controlled object process A second subtraction processing unit that outputs a feedback amount, a condition signal input unit that outputs a control condition at the time of operation as a control condition signal, and a typical one assuming a change in characteristics of a controlled process due to a change in the control condition. A filter time constant representative storage unit that stores a first time constant and a second time constant, a control condition signal output from the condition signal input unit, and a representative first stored in the filter time constant representative storage unit. And a filter time constant estimating unit for estimating and outputting the first time constant and the second time constant according to the control condition based on the time constant and the second time constant of IMC. controller. 3. The IMC controller according to claim 2, based on the representative first time constant and second time constant stored in the filter time constant representative storage unit, instead of the filter time constant estimation unit. A membership function in fuzzy inference is set, and the first time constant and the second time constant corresponding to the control condition are estimated and output based on the membership function and the control condition signal output from the condition signal input unit. An IMC controller having a filter time constant estimation processing unit for performing. 4. A reference control amount corresponding to a control amount of a control target process, which is a control result, is calculated by calculating an operation amount output to a control target process from a control target value, and using an internal model expressing the control target process by a mathematical expression. Control is performed by calculating and feeding back the difference between the control amount and the reference control amount I
In the MC controller, a target value filter unit that outputs the input target value with a characteristic determined by a first time constant, a first subtraction processing unit that subtracts a feedback amount from the output of the target value filter unit, A target value / disturbance filter unit that outputs the output of the first subtraction processing unit with a characteristic determined by a second time constant, and an operation amount from the output of the target value / disturbance filter unit based on the parameters of the internal model. An operation amount calculation unit including an operation unit that calculates a value, an internal model storage unit that stores parameters of the internal model, and a reference control based on an operation amount output from the operation amount calculation unit based on the parameters of the internal model. The internal model output calculation unit that calculates the quantity and the reference control amount output from the internal model output calculation unit from the control quantity of the controlled object process A second subtraction processing unit that outputs a feedback amount, a condition signal input unit that outputs a control condition signal during operation as a control condition signal, and a control condition signal that is output from the condition signal input unit. An IMC controller, comprising: a filter time constant calculation processing unit that calculates and outputs the first time constant and the second time constant.