JP7807944B2 - Parameter estimation apparatus and method - Google Patents

Description

The present invention relates to technology for estimating control device parameters, and in particular to a parameter estimation device and method for estimating parameters for an overshoot suppression function.

PID control, a typical type of feedback control, operates in the form of recovery control (disturbance recovery response) when a disturbance is applied to a temperature control system, and overshooting of the controlled variable often occurs due to factors such as the high heat retention of heating devices.

The heating device in Figure 10 is composed of a heat treatment furnace 100 that heats the workpiece to be treated, an electric heater 101, a temperature sensor 102 that measures the temperature inside the heat treatment furnace 100, a temperature controller 103 that controls the temperature inside the heat treatment furnace 100, a power regulator 104, a power supply circuit 105, and a PLC (Programmable Logic Controller) 106 that controls the entire heating device. The temperature controller 103 calculates a manipulated variable MV so that the temperature PV (controlled variable) measured by the temperature sensor 102 matches the temperature set point SP. The power regulator 104 determines the power corresponding to the manipulated variable MV and supplies this determined power to the electric heater 101 via the power supply circuit 105.

Figure 11 shows an example of an overshoot that occurred during temperature control of a heating device such as that shown in Figure 10. The example in Figure 11 shows an example in which a disturbance that causes the controlled variable PV to decrease occurs during temperature control, and when the manipulated variable MV corresponding to the disturbance is output from the temperature controller 103 and disturbance recovery control is performed, an overshoot occurs in which the controlled variable PV exceeds the temperature set value SP = 300°C due to excessive manipulated variable correction.

To address the phenomenon shown in Figure 11, an overshoot suppression function has been proposed (see Patent Document 1). The technology disclosed in Patent Document 1 calculates a correction amount Erx for the deviation Er (= SP - PV) between the set value SP and the controlled variable PV using the parameter ξ during the first half of the disturbance application period (when the temperature is decreasing in Figure 11), and calculates the correction amount Erx using the parameter λ during the second half of the disturbance application period (when the disturbance is converging).

Specifically, when the deviation Er is positive or negative and its absolute value |Er| increases, it is determined that a disturbance has been applied, and the deviation correction amount Erx is calculated as follows:
Erx=ξEr...(1)

The parameter ξ is, for example, 0.8. Subsequently, a convergence calculation is performed so that the deviation correction amount Erx gradually converges to 0 according to a fixed rule. Specifically, the deviation correction amount Erx' after the convergence calculation is calculated as follows:
Erx'=λErx...(2)

The parameter λ (0<λ<1) is, for example, λ = 0.95. By correcting the deviation Er using the deviation correction amount Erx' (Er - Erx'), overshoot is suppressed. The effect of the technology disclosed in Patent Document 1 is shown in Figure 12. Note that Figure 12 shows the corrected set value SPh = SP - Erx' instead of the deviation correction amount Erx'. As disclosed in Patent Document 1, the deviation correction Er - Erx' is equivalent to the set value correction (SP - Erx') - PV.

In the technology disclosed in Patent Document 1, the parameter λ is a constant. If an appropriate value for the parameter λ specific to the target control system can be determined, there is no need to frequently change the value. However, the conventional technology disclosed in Patent Document 1 had the problem of not being able to realize a technology for determining an appropriate parameter λ.

Patent No. 3437807

The present invention was made to solve the above-mentioned problems, and aims to provide a parameter estimation device and method that can efficiently determine and adjust appropriate parameters for the overshoot suppression function of a control device.

a parameter obtaining unit configured to obtain, from the control device, a parameter ξ to be multiplied by the deviation in order to calculate a deviation correction amount; an optimum value obtaining unit configured to calculate, based on a ratio between the maximum value of the overshoot amount and the maximum value of the disturbance deviation, the parameter ξ, and the time difference, an optimum value of a parameter λ that defines a time constant T for asymptotically converging the deviation correction amount to zero after the disturbance is applied; and an optimum value output unit configured to output the optimum value of the parameter λ.

In one configuration example of the parameter estimation device of the present invention, the optimum value calculation unit estimates an amount of overshoot of the controlled variable with respect to the set value after the control device has performed a set value correction equivalent to the deviation correction, based on the ratio between the maximum value of the overshoot amount and the maximum value of the disturbance deviation and the parameter ξ, calculates an optimum value of the time constant T so that the amount of deviation correction at a time point when the time difference has elapsed since the time point when the maximum value of the disturbance deviation was detected coincides with the estimated amount of overshoot, and converts this optimum value of the time constant T into an optimum value of the parameter λ.
In one example configuration of the parameter estimation device of the present invention, the optimum value calculation unit converts the optimum value of the time constant T into the optimum value of the parameter λ by λ=T/(T+dt), where dt is the control period of the control device.

Furthermore, the present invention provides a parameter estimation method for estimating parameters of a control device by a computer having a CPU and a storage device, wherein, in a state in which a function of the control device to correct a deviation between a set value and a controlled variable to suppress an overshoot of the controlled variable is disabled, the CPU is caused to execute the following steps according to a program stored in the storage device: a first step of detecting a maximum value of an overshoot amount of the controlled variable generated after a disturbance is applied to the controlled variable; a second step of detecting a maximum value of a disturbance deviation when the disturbance is applied; a third step of detecting a time difference from a point in time when the maximum value of the disturbance deviation is detected to a point in time when the maximum value of the overshoot amount is detected; a fourth step of acquiring from the control device a parameter ξ to be multiplied by the deviation in order to calculate a deviation correction amount; a fifth step of calculating, based on a ratio between the maximum value of the overshoot amount and the maximum value of the disturbance deviation, the parameter ξ, and the time difference, an optimum value of a parameter λ that defines a time constant T for asymptotically converging the deviation correction amount to 0 after the disturbance is applied; and a sixth step of outputting the optimum value of the parameter λ.

In one configuration example of the parameter estimation method of the present invention, the fifth step includes a step of estimating an amount of overshoot of the controlled variable with respect to the set value after a set value correction equivalent to the deviation correction has been performed in the control device, based on the ratio between the maximum value of the overshoot amount and the maximum value of the disturbance deviation and the parameter ξ, calculating an appropriate value of the time constant T so that the amount of deviation correction at a time point when the time difference has elapsed since the time point when the maximum value of the disturbance deviation was detected coincides with the estimated amount of overshoot, and converting this appropriate value of the time constant T into an appropriate value of the parameter λ.
In one configuration example of the parameter estimation method of the present invention, the fifth step is characterized by including a step of converting an appropriate value of the time constant T into an appropriate value of the parameter λ by λ=T/(T+dt), where dt is a control period of the control device.

According to the present invention, by providing an overshoot amount detection unit, a disturbance deviation detection unit, a response time difference detection unit, a parameter acquisition unit, an optimum value calculation unit, and an optimum value output unit, it is possible to efficiently determine and adjust the appropriate parameter λ for the overshoot suppression function of the control device.

FIG. 1 is a block diagram showing the configuration of a parameter estimation device according to an embodiment of the present invention. FIG. 2 is a flowchart illustrating the operation of the control device according to the embodiment of the present invention. FIG. 3 is a flowchart illustrating the operation of the parameter estimation device according to the embodiment of the present invention. FIG. 4 is a diagram showing the simulation results of the disturbance recovery response by a control device to which the conventional technology is applied. FIG. 5 is a diagram showing the simulation results of the disturbance recovery response of a conventional control device in which the overshoot suppression function is disabled. FIG. 6 is a diagram showing the simulation results of the disturbance recovery response of the control device after the parameters have been adjusted to appropriate values. FIG. 7 is a diagram showing the simulation results of the disturbance recovery response of a conventional control device in which the overshoot suppression function is disabled. FIG. 8 is a diagram showing the simulation results of the disturbance recovery response of the control device after the parameters have been adjusted to appropriate values. FIG. 9 is a block diagram showing an example of the configuration of a computer that realizes a parameter estimation device and a control device according to an embodiment of the present invention. FIG. 10 is a block diagram showing the configuration of the heating device. FIG. 11 is a diagram showing an example of an overshoot that occurs during temperature control of a heating device. FIG. 12 is a diagram showing the effect of the conventional overshoot suppression function.

[Principles of the Invention]
In the following explanation, the principle of the invention will be explained with the understanding that the set value SP is corrected. Furthermore, the parameter λ is a value that defines the rate at which the deviation correction amount Erx converges. However, since it is a ratio used in repeated calculations for each control cycle, it is easier to understand if it is replaced with the time constant T, which is an expression unrelated to the control cycle. Therefore, in the following explanation, the parameter λ will be replaced with the time constant T.

When considered as an operation to correct the set value SP, the operation of the technology disclosed in Patent Document 1 is equivalent to an operation in which the corrected set value SPg (= SP - Erx) also decreases at a ratio defined by the parameter ξ in sync with the decrease in the controlled variable PV, and then the set value SPg asymptotically converges to the original set value SP at a speed of time constant T. At this time, the controlled variable PV overshoots the corrected set value SPg (pseudo-overshoot), but it is preferable to correct the set value SP so that the controlled variable PV of this pseudo-overshoot approaches the original set value SP.

Therefore, a disturbance recovery response is attempted at least once, and the pseudo-overshoot amount OSx corresponding to the corrected set value SPg defined by the parameter ξ is calculated based on the ratio (overshoot rate) between the maximum overshoot amount OS_max and the maximum deviation Er_max due to the disturbance. Then, a time constant T equivalent to the parameter λ can be determined based on the amount of pseudo-overshoot and its occurrence timing.

In this invention, we assume that, for example, after parameter ξ has been appropriately adjusted, there may be a change in the nature of the controlled object or disturbance related to parameter λ. However, it is also possible to adjust parameter ξ by trial and error after adjusting parameter λ.

[Example]
An embodiment of the present invention will now be described with reference to the drawings. Fig. 1 is a block diagram showing the configuration of a parameter estimation device according to an embodiment of the present invention. The parameter estimation device 1 includes an overshoot amount detection unit 10 that detects a maximum value OS_max of an overshoot amount of the controlled variable PV that occurs after a disturbance is applied to the controlled variable PV when a function of a control device 2 to correct a deviation Er between a setpoint SP and a controlled variable PV to suppress an overshoot of the controlled variable PV is disabled, a disturbance deviation detection unit 11 that detects a maximum value Er_max of the disturbance deviation when the disturbance is applied, and a time interval between the detection of the maximum value Er_max of the disturbance deviation and the detection of the maximum value OS_max of the overshoot amount. a parameter acquiring unit 13 that acquires from the control device 2 a parameter ξ that is to be multiplied by the deviation Er to calculate the deviation correction amount Erx; an optimum value calculating unit 14 that calculates an optimum value of a parameter λ for converging the deviation correction amount Erx to 0 after the disturbance is applied, based on the ratio between the maximum value of the overshoot amount OS_max and the maximum value of the disturbance deviation Er_max, the parameter ξ, and the time difference tx; and an optimum value output unit 15 that outputs the optimum value of the parameter λ.

The control device 2 includes a control variable input unit 20 that receives a control variable PV as input, a set value input unit 21 that receives a set value SP as input, a subtraction unit 22 that calculates a deviation Er by subtracting the control variable PV from the set value SP, a disturbance application detection unit 23 that determines for each control cycle whether a disturbance has been applied based on the deviation Er and determines that a disturbance has been applied when the absolute value of the deviation Er increases, a deviation correction variable calculation unit 24 that calculates a deviation correction variable Erx based on the magnitude of the deviation Er when the disturbance application detection unit 23 detects that a disturbance has been applied, a deviation correction variable convergence calculation unit 25 that performs a convergence calculation so that the deviation correction variable Erx calculated by the deviation correction variable calculation unit 24 gradually converges to zero according to a fixed rule, a control calculation unit 26 that calculates a manipulated variable MV by PID control calculation based on the deviation Er and the deviation correction variable after the convergence calculation, and a manipulated variable output unit 27 that outputs the manipulated variable MV to the controlled object.

First, before explaining the parameter estimation device 1, we will briefly explain the operation of the control device 2 disclosed in Patent Document 1. Figure 2 is a flowchart explaining the operation of the control device 2. A control variable PV measured by a sensor (not shown) is input to the control variable input unit 20 (step S101 in Figure 2).

A set value SP set by an operator is input to the set value input unit 21 (step S102 in FIG. 2).
The subtraction unit 22 calculates the deviation Er between the set value SP and the controlled variable PV, ie, Er=SP-PV (step S103 in FIG. 2).

Next, the disturbance application detection unit 23 determines for each control cycle whether a disturbance has been applied, and outputs a start signal s1 to the deviation correction amount calculation unit 24 when it determines that a disturbance has been applied (step S104 in Figure 2). Specifically, the disturbance application detection unit 23 determines that a disturbance has been applied when the deviation Er is positive and the absolute value |Er| increases (when Er > 0 and Er > Er' (Er' is the deviation from one control cycle ago)), or when the deviation Er is negative and the absolute value |Er| increases (when Er < 0 and Er < Er').

When the start signal s1 is output from the disturbance application detection unit 23 (YES in step S104), the deviation correction amount calculation unit 24 calculates the deviation correction amount Erx by multiplying the deviation Er by the parameter ξ as shown in equation (1) and outputs this to the deviation correction amount convergence calculation unit 25 (step S105 in Figure 2).

Next, the deviation correction amount convergence calculation unit 25 calculates the deviation correction amount Erx' after the convergence calculation using equation (2) (step S106 in Figure 2). Note that the deviation correction amount calculation unit 24 outputs the deviation correction amount Erx only when the disturbance application detection unit 23 determines that a disturbance has been applied. In other words, when the situation in which the absolute value of the deviation |Er| is increasing (the situation in which the control amount PV deviates from the set value SP) ends, the output of the start signal s1 stops, and the output of the deviation correction amount Erx by the deviation correction amount calculation unit 24 stops.

When the deviation correction amount Erx is not output from the deviation correction amount calculation unit 24, the deviation correction amount convergence calculation unit 25 performs the convergence calculation of equation (2) using the deviation correction amount Erx' that was converged one control cycle ago as the deviation correction amount Erx for the current control cycle. The initial value of the deviation correction amount Erx' (Erx) is 0.

The control calculation unit 26 performs PID control calculation as shown in the following equation based on the value obtained by subtracting the deviation correction amount Erx' after the convergence calculation from the deviation Er, to calculate the manipulated variable MV (step S107 in FIG. 2).
MV=(100/Pb) {1+(1/Tis)+Tds}(Er-Erx')
...(3)

In equation (3), Pb is the proportional band, Ti is the integral time, Td is the differential time, and s is the Laplace operator.
The manipulated variable MV calculated by the control calculation unit 26 is output to the controlled object (actually, for example, the power regulator 104 in FIG. 10) via the manipulated variable output unit 27 (step S108 in FIG. 2).

The control device 2 executes steps S101 to S108 for each control cycle until control is terminated, for example, by an operator's instruction (YES in step S109 in Figure 2).

Next, the operation of the parameter estimation device 1 of this embodiment will be described with reference to Fig. 3. The parameter estimation device 1 is connected to the control device 2 when adjusting the parameter λ.
First, the overshoot amount detection unit 10 disables the overshoot suppression function of the control device 2 (step S200 in FIG. 3). When the overshoot suppression function is disabled, the deviation correction amount Erx′ becomes 0, and therefore the control calculation unit 26 of the control device 2 calculates the manipulated variable MV using the following equation (4) instead of equation (3).
MV=(100/Pb) {1+(1/Tis)+Tds}Er...(4)

After the disturbance application detection unit 23 of the control device 2 determines that a disturbance has been applied while the overshoot suppression function of the control device 2 is disabled (YES in step S201 in Figure 3), the overshoot amount detection unit 10 detects the maximum value OS_max of the overshoot amount of the controlled variable PV generated by the disturbance recovery response (the difference PV_max - SP between the maximum value PV_max of the increased controlled variable PV and the set value SP) (step S202 in Figure 3).

For example, in the case of the heating device shown in Figure 10, during the trial to determine the parameter λ, it is possible to intentionally create a disturbance that causes the temperature inside the heat treatment furnace 100 (controlled variable PV) to drop by temporarily opening the door of the heat treatment furnace 100.

The disturbance deviation detection unit 11 detects the maximum disturbance deviation Er_max (the difference SP - PV_min between the set value SP and the minimum value PV_min of the decreased controlled variable PV) when a disturbance is applied with the control device 2's overshoot suppression function disabled (step S203 in Figure 3).

The response time difference detection unit 12 detects the time difference tx from the time when the maximum value Er_max of the disturbance deviation is detected (the time when the decreased control amount PV reaches the minimum value PV_min) to the time when the maximum value OS_max of the overshoot amount is detected (the time when the increased control amount PV reaches the maximum value PV_max) (step S204 in Figure 3).
The parameter acquisition unit 13 acquires the parameter ξ set in the deviation correction amount calculation unit 24 of the control device 2 (step S205 in FIG. 3).

The optimum value calculation unit 14 calculates the corrected deviation Erh when the maximum value Er_max of the disturbance deviation (minimum value PV_min of the controlled variable PV) occurs, using the following equation, based on the maximum value Er_max of the disturbance deviation detected by the disturbance deviation detection unit 11 and the parameter ξ acquired by the parameter acquisition unit 13 (step S206 in FIG. 3 ):
Erh=Er_max-ξEr_max...(5)

Note that the corrected deviation Erh in this embodiment is not the corrected deviation Er_max-λξEr_max itself that the control device 2 uses when calculating the manipulated variable MV, but rather the value obtained by correcting the deviation Er_max using the deviation correction amount (Erx in equation (1)) before multiplying the second term on the right side of equation (5) by the parameter λ.

Next, since the corrected set value SPg can be estimated from the parameter ξ, the optimum value calculation unit 14 estimates the overshoot amount OSx of the controlled variable PV with respect to the set value SPg after setting value correction equivalent to the deviation correction, based on the parameter ξ, the maximum overshoot amount OS_max detected by the overshoot amount detection unit 10, and the maximum disturbance deviation Er_max detected by the disturbance deviation detection unit 11, using the following equation (step S207 in FIG. 3).
OSx=ErhOS_max/Er_max
=(Er_max-ξEr_max)OS_max/Er_max...(6)

Equation (6) is derived from the property that when there is no change in the characteristics of the PID control system of the control device 2, the ratio between the disturbance deviation and the amount of overshoot (overshoot rate) is maintained (Er_max:OS_max = Erh:OSx).

Note that the corrected set value SPg assumed in the calculation of equation (6) is not the corrected set value SPh = SP - Erx' itself, which corresponds to the corrected deviation used by the control device 2 when calculating the manipulated variable MV, but rather the value SPg = SP - Erx obtained by correcting the set value SP using the deviation correction amount Erx. Therefore, the deviation correction in the above "set value correction equivalent to deviation correction" does not refer to Er - Erx', but rather to Er - Erx. The deviation correction amount Erx = SP - SPg converges to 0.0 from ξEr_max, which corresponds to the initial value of the deviation correction amount, with the time constant T.

The elapsed time at which the maximum overshoot amount OSx is estimated to appear is the time difference tx. Therefore, the optimum value calculation unit 14 calculates an optimum value of the time constant T equivalent to the parameter λ by the following equation so that the deviation correction amount Erx at the elapsed time tx coincides with the overshoot amount OSx (step S208 in FIG. 3).
T=-tx/ln{OSx/(ξEr_max)} ...(7)

In equation (7), ln is the natural logarithm. Finally, the optimum value calculation unit 14 converts the optimum value of the time constant T into the optimum value of the parameter λ using the following equation (step S209 in FIG. 3). dt is the control period of the control device 2.
λ=T/(T+dt)...(8)

The optimum value output unit 15 outputs the optimum value of the parameter λ (step S210 in Figure 3). An example of the output is the display of the optimum value. The operator looks at the displayed optimum value and changes the parameter λ set in the control device 2. Alternatively, the optimum value output unit 15 may automatically update the parameter λ set in the deviation correction amount convergence calculation unit 25 of the control device 2 to the optimum value.

This completes the adjustment of parameter λ. After the adjustment is complete, the operator operates the parameter estimation device 1 to re-enable the overshoot suppression function of the control device 2 and disconnect the parameter estimation device 1 from the control device 2 (step S211 in Figure 3).

Figures 4 to 6 show a verification example of this embodiment when a change occurs in the controlled variable PV that has a strong tendency toward impulse disturbance (change in the manipulated variable MV ΔMV = 1.0%). Note that this verification example assumes that the parameter ξ of the control device 2 has been adjusted to a generally appropriate value of ξ = 0.95. However, after adjusting the parameter λ using the parameter estimation device 1, the parameter ξ may be fine-tuned by trial and error.

Figure 4 shows the results of a simulation of the disturbance recovery response of a control device 2 that applies the technology disclosed in Patent Document 1. In the example in Figure 4, the parameter ξ is set to a generally appropriate value of ξ = 0.95, but the time constant T equivalent to the parameter λ is set to an inappropriate value of T = 90 seconds. This results in excessive correction of the manipulated variable when a disturbance occurs that causes the controlled variable PV to decrease, resulting in an overshoot in which the controlled variable PV exceeds the set value SP = 300°C.

Here, we assume that the parameter ξ and time constant T were originally adjusted appropriately, but that a subsequent change in response speed caused the time constant T to become inappropriate. In other words, we assume that the object to be adjusted is the time constant T.

Figure 5 shows the simulation results of the disturbance recovery response by the control device 2 under the same conditions as Figure 4, but with the overshoot suppression function disclosed in Patent Document 1 disabled. In other words, the example in Figure 5 corresponds to the disturbance recovery response implemented in steps S200 to S204.

In this embodiment, the overshoot amount detection unit 10 detects the maximum overshoot amount OS_max = 8°C, the disturbance deviation detection unit 11 detects the maximum disturbance deviation Er_max = 20°C, and the response time difference detection unit 12 detects the time difference tx = 230 seconds. The parameter ξ set in the deviation correction amount calculation unit 24 of the control device 2 is 0.95. In this embodiment, the optimum value calculation unit 14 calculates the optimum value of the time constant T to be T = 102 seconds.

Figure 6 shows the simulation results of the disturbance recovery response of the control device 2 after adjusting the time constant T (parameter λ) to an appropriate value. In the example of Figure 6, the parameter ξ is set to a roughly appropriate value of ξ = 0.95, and the time constant T is set to an appropriate value of T = 102 seconds, which shows that overshooting of the controlled variable PV is appropriately suppressed.

Figures 7 and 8 show a verification example of this embodiment when a change in the controlled variable PV that has a strong tendency toward a step disturbance (change in the manipulated variable MV ΔMV = 30.0%) occurs. Note that this verification example assumes that the parameter ξ of the control device 2 has been adjusted to a roughly appropriate value of ξ = 0.50.

Figure 7 shows the simulation results of the disturbance recovery response by the control device 2 disclosed in Patent Document 1 with the overshoot suppression function disabled. In other words, the example in Figure 7 corresponds to the disturbance recovery response implemented in steps S200 to S204.

In this embodiment, the overshoot amount detection unit 10 detects the maximum overshoot amount OS_max = 5.5°C, the disturbance deviation detection unit 11 detects the maximum disturbance deviation Er_max = 23°C, and the response time difference detection unit 12 detects the time difference tx = 325 seconds. The parameter ξ set in the deviation correction amount calculation unit 24 of the control device 2 is 0.50. The optimum value calculation unit 14 in this embodiment calculates the optimum value of the time constant T to be T = 227 seconds.

Figure 8 shows the simulation results of the disturbance recovery response of the control device 2 after adjusting the time constant T (parameter λ) to an appropriate value. In the example of Figure 8, the parameter ξ is set to a roughly appropriate value of ξ = 0.50, and the time constant T is set to an appropriate value of T = 227 seconds, which shows that overshooting of the controlled variable PV is appropriately suppressed.

In this embodiment, the parameter estimation device 1 and the control device 2 are provided separately, but the parameter estimation device 1 may also be provided inside the control device 2.

The parameter estimation device 1 and control device 2 of this embodiment can each be realized by a computer equipped with a CPU (Central Processing Unit), a storage device, and an interface with the outside, and a program that controls these hardware resources. An example configuration of this computer is shown in Figure 9. The computer includes a CPU 200, a storage device 201, and an interface device (I/F) 202.

In the case of the parameter estimation device 1, the control device 2, a display device, etc. are connected to the I/F 202. In the case of the control device 2, the parameter estimation device 1, a sensor, a power regulator, etc. are connected to the I/F 202. A program for implementing the parameter estimation method of the present invention is stored in the storage device 201. The CPU 200 of each device executes the processing described in this embodiment in accordance with the program stored in the storage device 201.

This invention can be applied to technology for estimating control device parameters.

1...parameter estimation device, 2...control device, 10...overshoot amount detection unit, 11...disturbance deviation detection unit, 12...response time difference detection unit, 13...parameter acquisition unit, 14...optimum value calculation unit, 15...optimum value output unit, 20...control amount input unit, 21...set value input unit, 22...subtraction unit, 23...disturbance application detection unit, 24...deviation correction amount calculation unit, 25...deviation correction amount convergence calculation unit, 26...control calculation unit, 27...operation amount output unit.

Claims (6)

an overshoot amount detection unit configured to detect a maximum value of an overshoot amount of the controlled variable that occurs after a disturbance is applied to the controlled variable, when a function of the control device to correct a deviation between a set value and a controlled variable to suppress an overshoot of the controlled variable is disabled;
a disturbance deviation detection unit configured to detect a maximum value of the disturbance deviation when the disturbance is applied;
a response time difference detection unit configured to detect a time difference from a time point at which the maximum value of the disturbance deviation is detected to a time point at which the maximum value of the overshoot amount is detected;
a parameter acquisition unit configured to acquire, from the control device, a parameter ξ to be multiplied by the deviation in order to calculate a deviation correction amount;
an optimum value calculation unit configured to calculate an optimum value of a parameter λ that defines a time constant T for asymptotically converging the deviation correction amount to 0 after the disturbance is applied, based on a ratio between the maximum value of the overshoot amount and the maximum value of the disturbance deviation, the parameter ξ, and the time difference ;
and an optimum value output unit configured to output an optimum value of the parameter λ. 2. The parameter estimation device according to claim 1,
the optimum value calculation unit estimates an amount of overshoot of the controlled variable with respect to the set value after a set value correction equivalent to the deviation correction is performed in the control device, based on the ratio between the maximum value of the overshoot amount and the maximum value of the disturbance deviation and the parameter ξ, calculates an optimum value of the time constant T so that the amount of deviation correction at a time point when the time difference has elapsed since the time point when the maximum value of the disturbance deviation was detected coincides with the estimated amount of overshoot, and converts this optimum value of the time constant T into an optimum value of the parameter λ. 3. The parameter estimation device according to claim 2,
wherein the optimum value calculation unit converts the optimum value of the time constant T into the optimum value of the parameter λ by λ = T/(T + dt), where dt is a control period of the control device. A parameter estimation method for estimating parameters of a control device using a computer having a CPU and a storage device, comprising:
a first step of detecting a maximum value of an overshoot amount of the controlled variable that occurs after a disturbance is applied to the controlled variable, in a state in which a function of the control device to correct a deviation between a set value and a controlled variable to suppress an overshoot of the controlled variable is disabled;
a second step of detecting a maximum value of the disturbance deviation when the disturbance is applied;
a third step of detecting a time difference between a time point when the maximum value of the disturbance deviation is detected and a time point when the maximum value of the overshoot amount is detected;
a fourth step of acquiring, from the control device, a parameter ξ to be multiplied by the deviation in order to calculate a deviation correction amount;
a fifth step of calculating an appropriate value of a parameter λ that defines a time constant T for asymptotically converging the deviation correction amount to 0 after the disturbance is applied, based on a ratio between the maximum value of the overshoot amount and the maximum value of the disturbance deviation, the parameter ξ, and the time difference ;
a sixth step of outputting an appropriate value of the parameter λ ;
A parameter estimation method characterized by causing the CPU to execute the program stored in the storage device . 5. The parameter estimation method according to claim 4,
the fifth step includes a step of estimating an amount of overshoot of the controlled variable with respect to the set value after a set value correction equivalent to the deviation correction is performed in the control device, based on the ratio of the maximum value of the overshoot amount to the maximum value of the disturbance deviation and the parameter ξ, calculating an appropriate value of the time constant T so that the amount of deviation correction at a time point when the time difference has elapsed since the time point when the maximum value of the disturbance deviation was detected coincides with the estimated amount of overshoot, and converting this appropriate value of the time constant T into an appropriate value of the parameter λ. 6. The parameter estimation method according to claim 5,
the fifth step includes a step of converting the optimum value of the time constant T into an optimum value of the parameter λ by λ=T/(T+dt), where dt is a control period of the control device.