JP7521712B1 - Control device, control method, and program - Google Patents

Abstract

[Problem] To make a controlled variable of a controlled object follow a target value even in the presence of an unknown disturbance while effectively utilizing an existing controller. [Solution] A control device according to one aspect is disposed between a higher-level control device that outputs a first target value for a controlled variable of a controlled object and a lower-level control device that controls the controlled object according to a given second target value, and includes a model parameter estimation unit that estimates parameters of a closed-loop response model that models a response of a closed-loop control formed by the lower-level control device and the controlled object based on the controlled variable and the second target value, a target value conversion unit that calculates a third target value by converting the first target value so that the controlled variable asymptotically approaches the first controlled variable based on the first target value, the second target value, and the parameters, and a switching unit that outputs either the first target value or the third target value to the lower-level control device as the next second target value. [Selected Figure] Figure 2

Description

This disclosure relates to a control device, a control method, and a program.

Various control methods, such as PID (Proportional Integral Differential) control and model predictive control, are known as control methods aimed at making the controlled variable of a controlled object follow a target value. In addition, a technology called a reference governor is known as a technology related to these control methods (Non-Patent Document 1). A reference governor is a technology that shapes a target value in advance so that the manipulated variable or state variable of a control system satisfies a specified constraint condition.

Takeshi Hatanaka, Joji Takaba, "Integrated Design of Reference Governor and Preview Control", Transactions of the Institute of Systems, Control and Information Engineers, Vol.18, No.1, pp.39-41, 2005.

When controlling a control target using an existing controller that uses PID control, etc., unknown disturbances can cause the controlled variable to deviate from the target value. To address this issue, it is possible to introduce advanced controllers that use model predictive control, etc., but this would be very costly. For this reason, there is a demand for technology that can effectively utilize existing controllers while allowing the controlled variable to track the target value even when unknown disturbances are present.

This disclosure has been made in consideration of the above points, and aims to provide a technology that can make effective use of existing controllers and enable the control amount of a controlled object to track a target value even when an unknown disturbance is present.

A control device according to one aspect of the present disclosure is a control device disposed between a higher-level control device that outputs a first target value for a control amount of a controlled object and a lower-level control device that controls the controlled object according to a given second target value, and includes a model parameter estimation unit that estimates parameters of a closed-loop response model that models the response of a closed-loop control consisting of the lower-level control device and the controlled object based on the control amount and the second target value, a target value conversion unit that calculates a third target value by converting the first target value so that the controlled amount approaches the first target value based on the first target value, the second target value, and the parameters, and a switching unit that outputs either the first target value or the third target value to the lower-level control device as the next second target value.

While effectively utilizing existing controllers, it is possible to make the controlled variable of the controlled object track the target value even when unknown disturbances are present.

FIG. 2 is a diagram illustrating an example of a hardware configuration of a control device according to the first embodiment. FIG. 2 is a diagram illustrating an example of a functional configuration of a control device according to the first embodiment. 4 is a diagram illustrating an example of a detailed functional configuration of a target value conversion unit according to the first embodiment; FIG. FIG. 13 is a diagram for explaining an example of the operation of a closed loop response function. FIG. 11 is a diagram illustrating an example of a functional configuration of a control device according to a second embodiment. FIG. 11 is a diagram illustrating an example of a detailed functional configuration of a target value conversion unit according to a second embodiment. 11 is a diagram for explaining an example of the operation of a high-speed complement control unit; FIG. FIG. 1 is a diagram illustrating an example of a configuration of a conventional control system. FIG. 1 is a diagram illustrating an example of a configuration of a control system including a control device according to an embodiment. FIG. 1 is a diagram illustrating a configuration of a control system according to an embodiment. FIG. 13 is a diagram (part 1) showing the implementation results when only a PID controller is used. FIG. 11 is a diagram (part 1) showing an implementation result when the control device according to the first embodiment is used. FIG. 11 is a diagram (part 1) showing an implementation result when the control device according to the second embodiment is used. FIG. 1 is a diagram showing the estimation results of model parameters (part 1). FIG. 2 is a diagram (part 2) showing the implementation results when only the PID controller is used. FIG. 11 is a diagram (part 2) showing an implementation result when the control device according to the first embodiment is used. FIG. 11 is a diagram (part 2) showing the implementation results when the control device according to the second embodiment is used. FIG. 13 is a diagram (part 2) showing the estimation results of model parameters.

Each embodiment of the present invention will be described in detail below with reference to the drawings. A control device 10 will be described below that can make the control amount of the control object follow a target value even when an unknown disturbance is present, while effectively utilizing an existing controller. Hereinafter, a control device that controls the control object using a predetermined control method (e.g. PID control, etc.) is assumed as the existing controller, and will be referred to as the "lower control device 20." In addition, hereinafter, a plant is assumed as the control object, and will be referred to as the "controlled plant 30." However, the control object is not limited to a plant, and any device, equipment, facility, etc. can be the control object.

The lower-level control device 20 calculates and outputs an operation amount u for the controlled plant 30 so that the control amount y of the controlled plant 30 follows a given target value r. As a result, the controlled plant 30 is controlled according to this operation amount u, and a new control amount y is obtained. This type of control is called "closed-loop control" or simply "closed loop". However, in general, in plant control, disturbances v occur due to factors such as changes in the environment, changes in equipment, devices, facilities, etc., and changes in raw materials, and the control amount y may deviate from the target value r due to the influence of the disturbances v. In other words, the control amount y of the controlled plant 30 is affected not only by the operation amount u but also by unknown disturbances v, so the control amount y may deviate from the target value r.

Note that the disturbance v may or may not be measurable by a measuring instrument. Even if the disturbance v can be measured, it may be possible to measure it only once or several times a day. For this reason, it is often difficult to suppress the influence of the unknown disturbance v by adjusting the parameters of the lower-level control device 20 (e.g., adjusting the parameters of the PID control). That is, for example, when suppressing the influence of the unknown disturbance v by adjusting the parameters of the lower-level control device 20, there is no guarantee that the influence of the disturbance v can be suppressed by adjusting the parameters once, and there is a possibility that the parameters will need to be readjusted many times. On the other hand, for example, when suppressing the influence of the disturbance v by introducing an advanced controller using model predictive control or the like instead of the lower-level control device 20, identification of the process model, modification of the controller and on-site wiring, adjustment tests, etc. are required, which requires a lot of costs. In addition, for example, additional costs are required when installing a new sensor or the like to measure the unknown disturbance v or increasing the measurement frequency.

Therefore, in the control device 10 according to each of the following embodiments, a given target value (hereinafter, this target value is referred to as the "initial target value r ₀ ") is converted using the control variable y of the controlled plant 30, and either this converted target value or the initial target value r ₀ is output to the lower-level control device 20. This makes it possible, for example, to suppress the influence of an unknown disturbance v without adjusting the parameters of the lower-level control device 20. Note that, hereinafter, the target value converted from the initial target value r ₀ is referred to as the "converted target value r _m ", and the target value output to the lower-level control device 20 is referred to as the "applied target value r ₁ ".

[First embodiment]
The first embodiment will be described below.

<Hardware configuration example of the control device 10 according to the first embodiment>
An example of a hardware configuration of the control device 10 according to the first embodiment will be described with reference to Fig. 1. Fig. 1 is a diagram showing an example of a hardware configuration of the control device 10 according to the first embodiment.

As shown in FIG. 1, the control device 10 according to the first embodiment includes an input device 101, a display device 102, an external I/F 103, a communication I/F 104, a processor 105, and a memory device 106. Each of these pieces of hardware is connected to each other so as to be able to communicate with each other via a bus 107.

The input device 101 is, for example, a touch panel or various physical buttons. The display device 102 is, for example, a display panel. Note that the control device 10 does not have to have at least one of the input device 101 and the display device 102.

The external I/F 103 is an interface with an external device such as a recording medium 103a. Examples of the recording medium 103a include a secure digital memory card (SD memory card) and a universal serial bus (USB) memory card.

The communication I/F 104 is an interface for connecting the control device 10 to a communication network. The processor 105 is, for example, various types of arithmetic devices such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit). The memory device 106 is, for example, various types of storage devices such as an SSD (Solid State Drive), RAM (Random Access Memory), ROM (Read Only Memory), and flash memory.

Note that the hardware configuration shown in FIG. 1 is an example, and the control device 10 may have other hardware configurations. For example, the control device 10 may have various hardware in addition to the hardware shown in the figure.

<Functional configuration example of the control device 10 according to the first embodiment>
An example of a functional configuration of the control device 10 according to the first embodiment will be described with reference to Fig. 2. Fig. 2 is a diagram showing an example of a functional configuration of the control device 10 according to the first embodiment.

2, the control device 10 according to the first embodiment includes a target value conversion unit 111, a model parameter estimation unit 112, a switch 113, and a timer 114. Each of these components is realized, for example, by a process in which one or more programs installed in the control device 10 are executed by the processor 105 or the like, or by dedicated hardware or the like. Here, an initial target value _r0 is given to the control device 10 according to the first embodiment from a higher-level control device (e.g., a process computer, an operator terminal, or the like).

Hereinafter, the time of continuous time is t, and the time t of the initial target value _r0 is expressed as _r0 (t). Similarly, for other variables, when the time t is to be specified, "(t)" is written after the variable. Note that the initial target value _r0 (t) may or may not be constant.

The timer 114 operates as an operation trigger for the target value conversion unit 111 and the model parameter estimation unit 112 every main control period _Tc . The main control period _Tc is a period in which the control device 10 outputs the application target value _r1 to the lower-level control device 20. The value of the main control period _Tc is set in advance.

The target value conversion unit 111 receives _an initial target value r ₀ (t) at the current time t, a controlled variable y(t) at the current time t, a target variable width w, a model parameter estimate θ _est (t), and an application target value r ₁ (t) at the current time t, and outputs a converted target value r _m (t) at the current time t for each main control period T c. The converted target value r _m is a target value obtained by converting the initial target value r ₀ so that a target deviation e ₀ described later approaches (asymptotically approaches) 0. The model parameter estimate θ _est is an estimate of a model parameter θ of a plant response function {S _θ (t)} estimated by the model parameter estimation unit 112. The plant response function {S _θ (t)} is a function representing a closed-loop response model that models the response of a closed-loop control configured by the lower-level control device 20 and the controlled plant 30.

The plant response function {S _θ (t)} includes a model parameter θ. Hereinafter, the plant response function {S _θ (t)} is referred to as a "closed-loop response function {S _θ (t)}." In addition, the model parameter θ set in the closed-loop response function {S _θ (t)} is also referred to as a "model parameter setting value θ."

The model parameter estimation unit 112 receives the controlled variable y(t) at the current time t and the applied target value _r1 (t) at the current time t, and outputs the model parameter estimate _θest (t) for each main control period _Tc . Note that the initial value of the model parameter θ may be a zero vector, a value representing a unit step response, or any other arbitrary initial value.

The switch 113 outputs either the initial target value r ₀ (t) at the current time t or the conversion target value r _m (t) at the current time t to the lower control device 20 as the application target value r ₁ (t+T _c ) at the next time t+T _c . It is possible to determine in various ways whether the initial target value r ₀ or the conversion target value r _m is to be output as the application target value r _{1. For example, either the initial target value r 0 or the conversion target value r m may be output as the application target value r 1 according} to a user's instruction, or the conversion target value r _m may be output as the application target value r ₁ if the target deviation e ₀ is equal to or greater than a certain value, or the initial target value r ₀ may be output as the application target value r 1 if the target deviation e ₀ is not equal to or greater than a certain value. In addition, for example, the conversion target value r _m may be output as the application target value r ₁ in a certain time period, and the initial target value r ₀ may be output as the application target value r ₁ in other time periods, or the conversion target value r _m may be output as the application target value r 1 if some condition is satisfied, or the initial target value r ₀ may be output as the application target value r ₁ if some condition is satisfied.

As described above, the control device 10 according to the first embodiment acquires the controlled variable y of the controlled plant 30 for each main control period _Tc , calculates the model parameter estimate value θ _est , converts the given initial target value r ₀ into the converted target value r _m , and outputs the initial target value r ₀ or the converted target value r _m to the lower-level control device 20 as the applied target value r ₁ at the next time. In other words, the control device 10 according to the first embodiment sequentially calculates the model parameter estimate value θ _est for each main control period _Tc , and sequentially outputs the applied target value r ₁ to the lower-level control device 20. Thereafter, the lower-level control device 20 calculates and outputs the manipulated variable u for the controlled plant 30 so that the controlled variable y of the controlled plant 30 follows the applied target value r _1. As a result, even if an unknown disturbance v exists, it is possible to suppress the influence of the disturbance v and prevent the controlled variable y from deviating from the target value r. Moreover, since there is no need to make any changes to the existing controller, the lower-level control device 20, it is possible to suppress the effects of the disturbance v at low cost and safely, compared to, for example, replacing it with an advanced controller that uses model predictive control, etc.

<<Example of detailed functional configuration of target value conversion unit 111 according to the first embodiment>>
A detailed functional configuration example of the target value conversion unit 111 according to the first embodiment will be described with reference to Fig. 3. Fig. 3 is a diagram illustrating an example of a detailed functional configuration of the target value conversion unit 111 according to the first embodiment.

As shown in FIG. 3, the target value conversion unit 111 according to the first embodiment includes a difference calculator 121, a control parameter calculation unit 122, a time difference calculator 123, a target deviation prediction unit 124, a target change amount calculation unit 125, an adder 126, and a conversion target value constraint unit 127.

The difference calculator 121 calculates the difference between the initial target value r ₀ (t) and the controlled variable y(t) as the target deviation e ₀ (t). That is, the difference calculator 121 calculates the target deviation e ₀ (t) by e ₀ (t) = r ₀ (t) - y(t).

The control parameter calculation unit 122 calculates a control gain _kI and a look-ahead length _Tp as control parameters based on a closed loop response function {S _θ (t)} where θ=θ _est (t). Note that the control parameter calculation unit 122 may calculate the control gain _kI and the look-ahead length _Tp by, for example, the method described in Reference 1.

Specifically, the control parameter calculation unit 122 may calculate the look-ahead length _Tp , for example, as follows.

Find T _p , where S _θ (T _p )=β×S _θ (T _max )
Here, β is a preset adjustment coefficient, and it is preferable that 0<β≦1. T _max is a final time that is preset as a sufficiently long value.

The control parameter calculator 122 may calculate the control gain _kI by, for example, _kI = α × 1/( _gP + ε), where α is a preset adjustment coefficient, and is preferably 0 < α ≦ 1. ε is a small non-negative value. _gP is the gain at the look-ahead length _Tp , that is, _gp = _Sθ ( _Tp ).

The time differencing calculator 123 calculates an applied target value change amount dr ₁ which represents the time difference of the applied target value r ₁ during one main control cycle T _c . That is, the time differencing calculator 123 calculates the applied target value change amount dr ₁ (t) by dr ₁ (t) = r ₁ (t) - r ₁ (t - T _c ). Note that the applied target value r ₁ (t) is a target value set in the low-level control device 20 at the current time t.

The target deviation prediction unit 124 calculates the corrected target deviation e*(t) based on a closed loop response function {S _θ (t)} where θ=θ _est (t), the target deviation e ₀ (t), a time series of past applied target value change amounts dr ₁ {dr ₁ (t)} (hereinafter referred to as the "applied target value change amount series {dr ₁ (t)}"), and the look-ahead length _Tp . Note that the target deviation prediction unit 124 may calculate the corrected target deviation e ^* (t) by the method described in Reference 1, for example, using the applied target value change amount series { ^dr ₁ (t)} instead of the operation change amount time series {du(t)} described in Reference 1.

Specifically, the target deviation prediction unit 124 may, for example, calculate a value that is predicted to change when the controlled variable changes after _Tp has elapsed from the current time t due to the past applied target value change amount dr ₁ as the look-ahead response correction value _{y n} (t), and then calculate a value obtained by correcting the target deviation e ₀ (t) with the look-ahead response correction value y n (t) as the corrected target deviation e ^* (t). The corrected target deviation e ^* (t) may be calculated, for example, by e ^* (t) = e ₀ (t) - y _n (t). The look-ahead response correction value y _n (t) may be calculated, for example, by the method described in Reference 2 or Reference 3.

The target change amount calculation unit 125 calculates the target change amount dr(t) based on the corrected target deviation e ^* (t) and the control gain _kI . The target change amount calculation unit 125 calculates the target change amount dr(t) by, for example, dr(t) = _kI × e ^* (t).

The adder 126 calculates a target value (hereinafter referred to as "main control unit target value r _a (t)") by adding the target change amount dr(t) to the application target value r ₁ (t). That is, the adder 126 calculates the main control unit target value r _a (t) by r _a (t) = r ₁ (t) + dr(t).

Based on the main control unit target value r _a (t), the initial target value r ₀ (t), and the target variable width w, the conversion target value constraint unit 127 imposes upper and lower limit constraints so that the absolute value of the difference between the main control unit target value r _a (t) and the initial target value r ₀ (t) falls within the target variable width w, and then calculates the constrained value as the conversion target value r _m (t). Specifically, the conversion target value constraint unit 127 calculates the conversion target value r _m (t) as follows.

ここで、目標可変幅ｗは当初目標値ｒ_０をどの程度変化させてもよいかを表す値であり、その値は予め設定される。なお、目標可変幅ｗは一定の値であってもよいし、時刻ｔに応じて変化させてもよい。

Here, the target variable width w is a value that indicates how much the initial target value _r0 can be changed, and the value is set in advance. Note that the target variable width w may be a constant value, or may be changed according to the time t.

<<Operation of the closed loop response function {S _θ (t)}>>
The operation of the closed loop response function {S _θ (t)} will be described with reference to Fig. 4. Fig. 4 is a diagram for explaining an example of the operation of the closed loop response function {S _θ (t)}.

As shown in FIG. 4, when a model parameter setting value θ and a time t are input, the closed-loop response function {S _θ (t)} outputs a unit step response S _θ (t) of the closed-loop response model at time t after the lapse of time t from the initial time 0.

<Closed-loop response model>
Since time t is expressed as t = k × _Tc , where k is an integer equal to or greater than 0, hereinafter, time t and integer k will be regarded as the same thing. This allows notations such as y(k) and _r1 (k) to be regarded as the same thing as y(t) and _r1 (t). Note that such an integer k is also called an index representing a sample time (which may also be called discrete time).

In this case, the ARMA model shown below can be used as a closed-loop response model that models the response of the closed-loop control consisting of the lower-level control device 20 and the controlled plant 30.

なお、ｙ（ｋ）は制御量、ｒ_１（ｋ）は適用目標値、ｋはサンプル時間のインデックスである。

Here, y(k) is the controlled variable, r ₁ (k) is the applied target value, and k is the index of the sample time.

Now, let the state vector ξ(k) be:

また、インデックスｋにおけるモデルパラメータ推定値θ_ｅｓｔ（ｋ）を以下とする。

Moreover, the model parameter estimate θ _est (k) at index k is defined as follows:

このとき、制御量ｙ（ｋ）は、ｙ（ｋ）＝ξ（ｋ）^τθ_ｅｓｔ（ｋ－１）により算出することができる。

In this case, the control amount y(k) can be calculated by y(k)=ξ(k) ^τ θ _est (k−1).

The model parameter θ can be estimated, for example, using the method described in Reference 1 (recursive least squares method including a forgetting factor).

The initial value θ ₀ =θ _est (0) of the model parameter θ may be a zero vector or may be a value representing the following unit step response.

なお、その他の任意の初期値をモデルパラメータθの初期値θ_０＝θ_ｅｓｔ（０）としてもよい。

Any other initial value may be used as the initial value θ ₀ =θ _est (0) of the model parameter θ.

[Second embodiment]
A second embodiment will be described below. In the second embodiment, a control period _Tf (hereinafter, referred to as a "high-speed control period _Tf ") shorter than the main control period _Tc is used to complement the target value, as in Reference 4.

The second embodiment will mainly be described with respect to the differences from the first embodiment, and a description of components that may be the same as those in the first embodiment will be omitted. In other words, components that are not specifically described in the second embodiment may be the same as those in the first embodiment.

<Functional configuration example of the control device 10 according to the second embodiment>
An example of a functional configuration of the control device 10 according to the second embodiment will be described with reference to Fig. 5. Fig. 5 is a diagram showing an example of a functional configuration of the control device 10 according to the second embodiment.

5, the control device 10 according to the second embodiment has a timer 116 in addition to the components described in the first embodiment. The timer 116 is realized, for example, by a process in which one or more programs installed in the control device 10 are executed by the processor 105 or the like, or by dedicated hardware or the like. Here, in the control device 10 according to the second embodiment, a control amount y(t) is obtained for each high-speed control period _Tf , and the control amount y(t) is input to a high-speed complementation control unit 128 described later.

The timer 116 functions as an operation trigger for calculating a high-speed complementary target value r _b , which will be described later, for each high-speed control period T _f . The high-speed control period T _f is expressed as nT _f =T _c , where n is a predetermined arbitrary natural number.

<<Example of detailed functional configuration of target value conversion unit 111 according to the second embodiment>>
A detailed functional configuration example of the target value conversion unit 111 according to the second embodiment will be described with reference to Fig. 6. Fig. 6 is a diagram showing an example of a detailed functional configuration of the target value conversion unit 111 according to the second embodiment.

As shown in FIG. 6, the target value conversion unit 111 according to the second embodiment includes a high-speed complementation control unit 128 in addition to the components described in the first embodiment.

The high-speed complementary control unit 128 outputs a high-speed complementary target value _rb based on the main control unit target value r _a (t) and the control amount y(t). The high-speed complementary target value _rb is a target value that complements between the main control unit target value r _a (t) and the main control unit target value r _a (t+T _c ) for each high-speed control period T _f . That is, the high-speed complementary control unit 128 outputs a high-speed complementary target value _rb (t) for each high-speed control period T _f . However, for a time t where t=n ₁ T _f and t=n ₂ T _c for two integers n ₁ and n ₂ , r _a (t)=r _b (t).

The conversion target value constraint unit 127 imposes upper and lower limit constraints based on the high speed complementary target value r _b (t), the initial target value r ₀ (t), and the target variable width w so that the absolute value of the difference between the high speed complementary target value r _b (t) and the initial target value r ₀ (t) falls within the target variable width w, and calculates the constrained value as the conversion target value r _m (t). That is, the conversion target value constraint unit 127 calculates the conversion target value r _m (t) for each high speed control period T _f by an equation in which r _a (t) is replaced with r _b (t) in the above equation 1. As a result, for example, a high speed complementary target value r _b that can quickly respond to changes in the controlled variable y and changes in the initial target value r ₀ is calculated for each high speed control period T _f .

<<Operation of the high-speed complement control unit 128>>
The operation of the high-speed complement control unit 128 according to the second embodiment will be described with reference to Fig. 7. Fig. 7 is a diagram for explaining an example of the operation of the high-speed complement control unit 128.

7, the main control unit target value r _a (t) changes for each main control period T _c , and the high-speed complementary target value r _b (t) calculated by the high-speed complementary control unit 128 according to the second embodiment changes for each high-speed control period T _f . However, for a time t where t=n ₁ T _f and t=n ₂ T _c for two integers n ₁ and n ₂ , r _a (t)=r _b (t) . This is because the high-speed complementary target value r _b is calculated based on the main control unit target value r _a at a timing when the target change amount calculation unit 125 is not operating within the main control period T _c .

The high speed complement control unit 128 may calculate the high speed complement target value r _b by, for example, a process similar to the process for calculating the complement operation amount u _b described in Reference 4. Specifically, the high speed complement control unit 128 may calculate the high speed complement target value r _b by, for example, the following steps 1 to 6. For simplicity, the following will describe a case in which the high speed complement target value r b (t') is calculated at time t _' , which is a continuous time that satisfies t+(n ₃ -1)T _{c ≦} t'<t+n ₃ T _c, for an integer n 3. In the following, an index that satisfies 0≦j<n+1 is used. Here, n is a natural number that satisfies nT _f =T _c .

Step 1: The high-speed completion control unit 128 determines whether j = n.

Step 2: If it is determined in step 1 that j=n, the high speed complement control unit 128 initializes _y0 =y(t), _mdr (0)=0, and j=0. Here, _mdr (j) is the average target value change amount at index j, as described later. Note that at this time, _the time t'=t is such that t= _{n1Tf and} t= _n2Tc for two integers _n1 and _n2 .

Step 3: If it is not determined in step 1 above that j = n, the high-speed completion control unit 128 updates the index so that j ← j + 1.

Step 4: The high speed complementation control unit 128 calculates the predicted target deviation e _f (j). The predicted target deviation e _f (j) is a provisional target deviation predicted at the current time t' after the look-ahead length T _p . The high speed complementation control unit 128 may calculate the predicted target deviation e _f (j) by, for example, the following calculation method 1 or calculation method 2.

Calculation method 1 of predicted target deviation e _f (j)
e _f (j)=r ₀ (t')-(y(t')+(r ₀ (t')-y ₀ )・(1-a(j)))
a(j)=j・T _f /T _p
Here, r ₀ (t') is the value of r ₀ (t) at t=t'.

Calculation method 2 of predicted target deviation e _f (j)
e _f (j) = r ₀ (t') - (y ₀ + (y (t') - y ₀ )/a (j))
a(j)=j・T _f /T _p

Step 5: The high speed complement control unit 128 calculates the average target value change m _dr (j) as follows.

dr _f (j)=k _I・e _f (j)
m _dr (j) = (m _dr (j-1)・j+dr _f (j))/(j+1)

Alternatively, to take a simple average, the average target value change m _dr (j) may be calculated as follows:

dr _f (j)=k _I・e _f (j)
m _dr (j)=m _dr (j-1)+dr _f (j)・T _f /T _p

Step 6: The high speed complement control unit 128 calculates the high speed complement target value r _b (t') by r _b (t') = r _a (t) + m _dr (j).

[Example of control system configuration]
An example of the configuration of a control system including the control device 10 according to the first and second embodiments will be described below.

<Example of a conventional control system configuration>
First, a configuration example of a conventional control system will be described with reference to Fig. 8. Fig. 8 is a diagram showing an example of the configuration of a conventional control system.

As shown in FIG. 8, in a conventional control system, a higher-level control device 40 (e.g., a process computer or an operator terminal) exists above a lower-level control device 20, and a target value r is given from the higher-level control device 40 to the lower-level control device 20. The lower-level control device 20 then calculates and outputs an operation amount u for the controlled plant 30 so that the controlled amount y of the controlled plant 30 follows the given target value r.

<Configuration example of a control system including the control device 10>
Next, a configuration example of a control system including the control device 10 according to the first or second embodiment will be described with reference to Fig. 9. Fig. 9 is a diagram showing an example of the configuration of a control system including the control device 10 according to one embodiment.

As shown in FIG. 9, in a control system including the control device 10 according to the first or second embodiment, the control device 10 is included as an intermediate control device between the upper control device 40 and the lower control device 20. In other words, the control device 10 is disposed as an intermediate control device between the upper control device 40 and the lower control device 20. In this control system, an initial target value r ₀ is given to the control device 10 from the upper control device 40, and the control device 10 calculates an application target value r ₁ based on the given initial target value r ₀ and the control amount y of the controlled plant 30, and outputs this application target value r ₁ to the lower control device 20. Thereafter, the lower control device 20 calculates and outputs an operation amount u for the controlled plant 30 so that the control amount y of the controlled plant 30 follows the given application target value r _1. Note that, in the example shown in FIG. 9, the control device 10 acquires the control amount y from the controlled plant 30, but the control device 10 may acquire the control amount y from, for example, the lower control device 20.

When incorporating the control device 10 according to the first or second embodiment into a conventional control system, all that is required is the installation of the control device 10 and the associated wiring of the transmission equipment, and there is no need to modify the upper control device 40 or the lower control device 20. Therefore, it is possible to incorporate the control device 10 according to the first or second embodiment into a conventional control system at low cost and safely.

The lower-level control device 20 may be any controller that controls the controlled plant 30 using a predetermined control method. For example, any controller that controls the controlled plant 30 using any existing control method such as PID control, model predictive control (MPC), sliding mode control, H2 control, etc. may be used as the lower-level control device 20.

In addition, as the control device 10 (intermediate control device), for example, a PC (personal computer), a distributed control system (DCS: Distributed Control System), a PLC (Programmable Logic Controller), a regulator, etc. can be used as the control device 10 (intermediate control device).

[Example]
An example of a control system including the control device 10 according to the first and second embodiments will be described below.

<System Configuration>
The configuration of the control system according to this embodiment will be described with reference to Fig. 10. Fig. 10 is a diagram showing the configuration of the control system according to one embodiment.

As shown in Fig. 10, in this embodiment, an initial target value _r0 is provided from a higher-level control device 40 such as a process computer to a control device 10 according to the first or second embodiment. The control device 10 according to the first or second embodiment calculates an application target value _r1 from the provided initial target value _r0 , and outputs it to a lower-level control device 20 which is a PID controller C. Then, the lower-level control device 20 calculates an operation amount u so that a control amount y of a controlled plant 30 which is a plant P follows the application target value _r1 , and outputs it to the controlled plant 30. In addition to the operation amount u, an unknown disturbance v is input to the controlled plant 30.

Transfer Function of Plant P The transfer function of the controlled plant 30 is a second-order lag system represented as follows.

ここで、ＧはプラントＰのゲイン、Ｔ_１及びＴ_２は時定数である。

where G is the gain of the plant P, and _T1 and _T2 are time constants.

PID controller C
The PID controller C is assumed to be a PI controller represented as follows:

ここで、ｋ_ｐは比例ゲイン、Ｔ_Ｉは積分ゲインである。なお、本実施例では微分成分を持たないが、微分成分を有するＰＩＤ制御器が用いられてもよい。また、ＰＩＤ制御器の構成として、微分先行型ＰＩＤ、比例微分先行型ＰＩＤ、速度型ＰＩＤ、２自由度ＰＩＤ等の各種のＰＩＤ制御が用いられてもよい。

Here, _kp is a proportional gain, and T _I is an integral gain. Although this embodiment does not have a differential component, a PID controller having a differential component may be used. In addition, various types of PID control such as a differential leading type PID, a proportional/differential leading type PID, a speed type PID, and a two-degree-of-freedom PID may be used as the configuration of the PID controller.

In addition, below, we consider the following two cases regarding PID parameters:

Case 1: When the PID parameters are well adjusted Case 2: When the PID parameters are not well adjusted Settings of the control device 10 The settings of the control device 10 according to the first or second embodiment were such that the main control period _Tc was _Tc = 5 [sec] and the target variable width w was w = 0.5. The settings of the control device 10 according to the second embodiment were such that the high-speed control period _Tf was _Tf = 0.5 [sec].

Under the above, in each of Case 1 and Case 2, the controlled plant 30 was controlled using only the PID controller C, using the control device 10 according to the first embodiment, and using the control device 10 according to the second embodiment. The implementation results are explained below.

<Results of Case 1>
<When only PID controller C is used>
In case 1, the implementation result when only the PID controller C (lower control device 20) is used will be described with reference to Fig. 11. Fig. 11 is a diagram (part 1) showing the implementation result when only the PID controller C is used.

Figure 11(a) shows a time series graph of the initial target value _r0 (t), Figure 11(b) shows the initial target value _r0 (t) and the controlled variable y(t), Figure 11(c) shows the disturbance v(t), and Figure 11(d) shows the manipulated variable u(t).

11(a) to 11(d), the initial target value r ₀ (t) is constant at 60, but the disturbance v(t) varies between -1 and 1. The disturbance v(t) causes the controlled variable y(t) to vary, with large fluctuations occurring especially at the beginning of the change in the disturbance v(t). In the PID control by the low-level control device 20, the controlled variable y(t) is made to follow the initial target value r ₀ (t), so the manipulated variable u(t) varies in the opposite direction to the disturbance v(t) so as to cancel out the disturbance v(t).

<<When the control device 10 according to the first embodiment is used>>
In case 1, the implementation result when the control device 10 according to the first embodiment is used will be described with reference to Fig. 12. Fig. 12 is a diagram (part 1) showing the implementation result when the control device 10 according to the first embodiment is used.

Fig. 12(a) shows a time series graph of the initial target value _r0 (t) and the converted target value _rm (t), Fig. 12(b) shows the initial target value _r0 (t) and the controlled variable y(t), Fig. 12(c) shows the disturbance v(t), and Fig. 12(d) shows the manipulated variable _u (t). The switch 113 is set to output the converted target value rm(t) as the applied target value _r1 (t+ _Tc ) at the next time t+ _Tc .

12(a) to 12(d), the conversion target value r _m (t) changes in response to changes in the disturbance v(t) and is a value different from the initial target value r ₀ (t), but the absolute value of the difference between the conversion target value r _m (t) and the initial target value r ₀ (t) is restricted to within a target variable width w = 0.5. Therefore, the conversion target value r _m (t) is a value between 59.5 and 60.5.

By setting the conversion target value r _m (t) in the low-level control device 20, the manipulated variable u(t) changes more sharply when the disturbance v(t) changes, compared to when only the PID controller C is used. As a result, the difference between the controlled variable y(t) and the initial target value r ₀ (t) is kept smaller than when only the PID controller C is used.

<<When the control device 10 according to the second embodiment is used>>
In case 1, the implementation result when the control device 10 according to the second embodiment is used will be described with reference to Fig. 13. Fig. 13 is a diagram (part 1) showing the implementation result when the control device 10 according to the second embodiment is used.

Fig. 13(a) shows a time series graph of the initial target value _r0 (t) and the converted target value _rm (t), Fig. 13(b) shows the initial target value _r0 (t) and the controlled variable y(t), Fig. 13(c) shows the disturbance v(t), and Fig. 13(d) shows the manipulated variable _u (t). The switch 113 is set to output the converted target value rm(t) as the applied target value _r1 (t+ _Tc ) at the next time t+ _Tc .

13(a) to 13(d), the conversion target value r _m (t) changes in response to the change in the disturbance v(t) and is a value different from the initial target value r ₀ (t), but the absolute value of the difference between the conversion target value r _m (t) and the initial target value r ₀ (t) is restricted to within a target variable width w = 0.5. Therefore, the conversion target value r _m (t) is a value between 59.5 and 60.5.

By setting the conversion target value r _m (t) in the low-level control device 20, the manipulated variable u(t) changes more sharply when the disturbance v(t) changes, compared to when only the PID controller C is used. As a result, the difference between the controlled variable y(t) and the initial target value r ₀ (t) is kept smaller than when only the PID controller C is used. Furthermore, the fluctuation of the controlled variable y(t) is further suppressed, compared to when the control device 10 according to the first embodiment is used.

<Comparison of control response>
In Case 1, the maximum and minimum values of the control response and the amount of overshoot in each of the cases where only the PID controller C is used, where the control device 10 according to the first embodiment is used, and where the control device 10 according to the second embodiment is used are shown in Table 1 below.

上記の表１に示すように、上側のオーバーシュート量は、ＰＩＤ制御のみを用いた場合では０．４２３、第一の実施形態に係る制御装置１０を用いた場合では０．３５２、第二の実施形態に係る制御装置１０を用いた場合では０．２８３である。また、下側のオーバーシュート量は、ＰＩＤ制御のみを用いた場合では０．８４６、第一の実施形態に係る制御装置１０を用いた場合では０．６７７、第二の実施形態に係る制御装置１０を用いた場合では０．５９０である。このため、第一又は第二の実施形態に係る制御装置１０を用いることにより、未知の外乱ｖの影響が抑制できていることがわかる。

As shown in Table 1 above, the amount of overshoot on the upper side is 0.423 when only PID control is used, 0.352 when the control device 10 according to the first embodiment is used, and 0.283 when the control device 10 according to the second embodiment is used. The amount of overshoot on the lower side is 0.846 when only PID control is used, 0.677 when the control device 10 according to the first embodiment is used, and 0.590 when the control device 10 according to the second embodiment is used. Therefore, it can be seen that the influence of the unknown disturbance v can be suppressed by using the control device 10 according to the first or second embodiment.

<Estimation result of model parameter θ>
In case 1, the estimation results of the model parameter θ when the control device 10 according to the first embodiment is used and when the control device 10 according to the second embodiment is used will be described with reference to Fig. 14. Fig. 14 is a diagram (part 1) showing the estimation results of the model parameter θ. Note that the ARMA model shown in the above formula 2 was used as the closed loop response model.

Figure 14(a) shows the estimation results of the model parameter θ when the control device 10 according to the first embodiment is used, and Figure 14(b) shows a time series graph of the estimation results of the model parameter θ when the control device 10 according to the second embodiment is used. It can be seen that the model parameter θ is estimated sequentially online in both cases where the control device 10 according to the first embodiment and the control device 10 according to the second embodiment are used.

<Results of Case 2>
<When only PID controller C is used>
In case 2, the implementation result when only the PID controller C (lower control device 20) is used will be described with reference to Fig. 15. Fig. 15 is a diagram (part 2) showing the implementation result when only the PID controller C is used.

Figure 15(a) shows a time series graph of the initial target value _r0 (t), Figure 15(b) shows the initial target value _r0 (t) and the controlled variable y(t), Figure 15(c) shows the disturbance v(t), and Figure 15(d) shows the manipulated variable u(t).

15(a) to 15(d), the initial target value r ₀ (t) is constant at 60, but the disturbance v(t) varies between -1 and 1. The disturbance v(t) causes the controlled variable y(t) to vary, with large fluctuations occurring especially at the beginning of the change in the disturbance v(t). In the PID control by the low-level control device 20, the controlled variable y(t) is made to follow the initial target value r ₀ (t), so the manipulated variable u(t) varies in the opposite direction to the disturbance v(t) so as to cancel out the disturbance v(t).

<<When the control device 10 according to the first embodiment is used>>
In case 2, the implementation result when the control device 10 according to the first embodiment is used will be described with reference to Fig. 16. Fig. 16 is a diagram (part 2) showing the implementation result when the control device 10 according to the first embodiment is used.

Fig. 16(a) shows a time series graph of the initial target value _r0 (t) and the converted target value _rm (t), Fig. 16(b) shows the initial target value _r0 (t) and the controlled variable y(t), Fig. 16(c) shows the disturbance v(t), and Fig. 16(d) shows the manipulated variable _u (t). The switch 113 is set to output the converted target value rm(t) as the applied target value _r1 (t+ _Tc ) at the next time t+ _Tc .

16(a) to 16(d), the conversion target value r _m (t) changes in response to the change in the disturbance v(t) and is a value different from the initial target value r ₀ (t), but the absolute value of the difference between the conversion target value r _m (t) and the initial target value r ₀ (t) is restricted to within a target variable width w = 0.5. Therefore, the conversion target value r _m (t) is a value between 59.5 and 60.5.

By setting the conversion target value r _m (t) in the low-level control device 20, the manipulated variable u(t) changes more sharply when the disturbance v(t) changes, compared to when only the PID controller C is used. As a result, the difference between the controlled variable y(t) and the initial target value r ₀ (t) is kept smaller than when only the PID controller C is used.

<<When the control device 10 according to the second embodiment is used>>
In case 2, the implementation result when the control device 10 according to the second embodiment is used will be described with reference to Fig. 17. Fig. 17 is a diagram (part 2) showing the implementation result when the control device 10 according to the second embodiment is used.

Fig. 17(a) shows a time series graph of the initial target value _r0 (t) and the converted target value _rm (t), Fig. 17(b) shows the initial target value _r0 (t) and the controlled variable y(t), Fig. 17(c) shows the disturbance v(t), and Fig. 17(d) shows the manipulated variable _u (t). The switch 113 is set to output the converted target value rm(t) as the applied target value _r1 (t+ _Tc ) at the next time t+ _Tc .

17(a) to 17(d), the conversion target value r _m (t) changes in response to the change in the disturbance v(t) and is a value different from the initial target value r ₀ (t), but the absolute value of the difference between the conversion target value r _m (t) and the initial target value r ₀ (t) is restricted to within a target variable width w = 0.5. Therefore, the conversion target value r _m (t) is a value between 59.5 and 60.5.

By setting the conversion target value r _m (t) in the low-level control device 20, the manipulated variable u(t) changes more sharply when the disturbance v(t) changes, compared to when only the PID controller C is used. As a result, the difference between the controlled variable y(t) and the initial target value r ₀ (t) is kept smaller than when only the PID controller C is used. Furthermore, the fluctuation of the controlled variable y(t) is further suppressed, compared to when the control device 10 according to the first embodiment is used.

<Comparison of control response>
In Case 2, the maximum and minimum values of the control response and the amount of overshoot in each of the cases where only the PID controller C is used, where the control device 10 according to the first embodiment is used, and where the control device 10 according to the second embodiment is used are shown in Table 2 below.

上記の表２に示すように、上側のオーバーシュート量は、ＰＩＤ制御のみを用いた場合では０．３９５、第一の実施形態に係る制御装置１０を用いた場合では０．３７９、第二の実施形態に係る制御装置１０を用いた場合では０．２５９である。また、下側のオーバーシュート量は、ＰＩＤ制御のみを用いた場合では０．７８９、第一の実施形態に係る制御装置１０を用いた場合では０．６３３、第二の実施形態に係る制御装置１０を用いた場合では０．５４７である。このため、第一又は第二の実施形態に係る制御装置１０を用いることにより、未知の外乱ｖの影響が抑制できていることがわかる。

As shown in Table 2 above, the amount of overshoot on the upper side is 0.395 when only PID control is used, 0.379 when the control device 10 according to the first embodiment is used, and 0.259 when the control device 10 according to the second embodiment is used. The amount of overshoot on the lower side is 0.789 when only PID control is used, 0.633 when the control device 10 according to the first embodiment is used, and 0.547 when the control device 10 according to the second embodiment is used. Therefore, it can be seen that the influence of the unknown disturbance v can be suppressed by using the control device 10 according to the first or second embodiment.

<Estimation result of model parameter θ>
In case 2, the estimation results of the model parameter θ when the control device 10 according to the first embodiment is used and when the control device 10 according to the second embodiment is used will be described with reference to Fig. 18. Fig. 18 is a diagram (part 2) showing the estimation results of the model parameter θ. Note that the ARMA model shown in the above formula 2 was used as the closed loop response model.

Figure 18(a) shows the estimation results of the model parameter θ when the control device 10 according to the first embodiment is used, and Figure 18(b) shows a time series graph of the estimation results of the model parameter θ when the control device 10 according to the second embodiment is used. It can be seen that the model parameter θ is estimated sequentially online in both cases where the control device 10 according to the first embodiment and the control device 10 according to the second embodiment are used.

[summary]
As described above, the control device 10 according to the first and second embodiments makes it possible to make the control amount of the controlled object follow the target value even when an unknown disturbance exists, while effectively utilizing an existing controller. In this case, the control device 10 according to the first and second embodiments does not require any modification to the existing controller, so that the target deviation against the unknown disturbance can be improved at low cost and safely. This is one of the notable features not found in the conventionally known methods for improving the target deviation (e.g., a method for readjusting PID parameters, a method for replacing with advanced control such as model predictive control, a method for modifying a closed loop such as a disturbance observer or Smith compensator, etc.).

In addition, after the control device 10 according to the first and second embodiments is introduced into the control system, even if, for example, the lower-level control device 20 is readjusted or the controlled plant 30 itself is changed, the model parameter θ is adjusted autonomously by the model parameter estimation unit 112, so there is also a feature that maintenance of the control device 10 (intermediate control device) itself is not required.

Furthermore, the control device 10 according to the first and second embodiments maintains the original control loop and constrains the converted target value r _m so that it does not deviate from the initial target value r ₀ by the target variable width w. Therefore, if the original control loop is stable, the stability of the control system incorporating the control device 10 is also guaranteed.

In addition, by using the control device 10 according to the first and second embodiments, it is possible to suppress the influence of the unknown disturbance v without installing a new sensor or the like for measuring the unknown disturbance v or increasing the frequency of measurement.

The present invention is not limited to the above-mentioned specifically disclosed embodiments, and various modifications, changes, and combinations with known technologies are possible without departing from the scope of the claims.

[References]
Reference 1: Japanese Patent No. 7014330 Reference 2: International Publication No. 2016/092872 Reference 3: Japanese Patent Publication No. 2020-21411 Reference 4: Japanese Patent No. 7283646

REFERENCE SIGNS LIST 10 Control device 20 Lower level control device 30 Control target plant 101 Input device 102 Display device 103 External I/F
103a Recording medium 104 Communication I/F
105 Processor 106 Memory device 107 Bus 111 Target value conversion section 112 Model parameter estimation section 113 Switch 114 Timer 115 Timer 121 Differentiator 122 Control parameter calculation section 123 Time differentiator 124 Target deviation prediction section 125 Target change amount calculation section 126 Adder 127 Conversion target value constraint section 128 High-speed complement control section

Claims (9)

A control device disposed between a higher-level control device that outputs a first target value for a controlled variable of a controlled object and a lower-level control device that controls the controlled object in accordance with a given second target value,
a model parameter estimation unit that estimates parameters of a closed-loop response model that models a response of a closed-loop control configured with the lower-level control device and the controlled object based on the controlled variable and the second target value;
a target value conversion unit that calculates a third target value by converting the first target value based on the first target value, the second target value, and the parameter so that the controlled variable asymptotically approaches the first target value;
a switching unit that outputs either the first target value or the third target value to the lower-level control device as the next second target value;
A control device having the above configuration. The control device according to claim 1, wherein the third target value is a value constrained so that the absolute value of the difference between the third target value and the first target value is equal to or less than a predetermined target variable range. The target value conversion unit includes:
a target deviation prediction unit that predicts a corrected target deviation obtained by correcting a target deviation representing a difference between the first target value and the controlled variable, based on the parameter, the first target value, the second target value, and the controlled variable, using the controlled variable after a predetermined look-ahead length;
a target change amount calculation unit that calculates a target change amount that represents a change amount of the second target value based on the corrected target deviation;
an adder that calculates a fourth target value by adding the target change amount to the second target value;
3. The control device according to claim 2, further comprising: a conversion target value constraint unit that calculates, as the third target value, the fourth target value constrained so that an absolute value of a difference between the fourth target value and the first target value is equal to or less than the target variable width. The target value conversion unit
4. The control device according to claim 3, further comprising: a control unit for calculating the third target value by converting the first target value in a predetermined first control cycle so that the controlled variable asymptotically approaches the first target value. The target value conversion unit includes:
5. The control device according to claim 4, further comprising a high-speed complementation control unit that calculates, in a predetermined second control period shorter than the first control period, a fifth target value that complements between the fourth target value and the fourth target value after the first control period, so that the controlled variable asymptotically approaches the first target value. the closed loop response model is an ARMA model;
The model parameter estimation unit
A control device according to claim 1 , further comprising a step of recursively estimating the parameters of the ARMA model. The upper control device is a process computer or an operator terminal,
The control device according to claim 1 , wherein the lower level control device is a DCS, a PLC, or a controller. A control device disposed between a higher-level control device that outputs a first target value for a controlled variable of a controlled object and a lower-level control device that controls the controlled object in accordance with a given second target value,
a model parameter estimation step of estimating parameters of a closed-loop response model that models a response of a closed-loop control configured with the lower-level control device and the controlled object based on the controlled variable and the second target value;
a target value conversion step of calculating a third target value by converting the first target value based on the first target value, the second target value, and the parameter so that the controlled variable asymptotically approaches the first target value;
a switching procedure of outputting either the first target value or the third target value to the lower-level control device as the next second target value;
A control method for performing the above. A control device disposed between a higher-level control device that outputs a first target value for a controlled variable of a controlled object and a lower-level control device that controls the controlled object in accordance with a given second target value,
a model parameter estimation step of estimating parameters of a closed-loop response model that models a response of a closed-loop control configured with the lower-level control device and the controlled object based on the controlled variable and the second target value;
a target value conversion step of calculating a third target value by converting the first target value based on the first target value, the second target value, and the parameter so that the controlled variable asymptotically approaches the first target value;
a switching procedure of outputting either the first target value or the third target value to the lower-level control device as the next second target value;
A program that executes the following.

Images