Two-freedom degree controller

Description: TECHNICAL FIELD [Detailed description of the invention]

【0001】

[Technical field to which the invention belongs]

The present invention relates to a two-degree-of-freedom control device used in various plant instrumentation fields such as petrochemical, chemical, steel, electric power, and papermaking, and in a servo control field, and in particular, the flow rate, pressure, liquid level, and composition of each part of the plant. It relates to a two-degree-of-freedom controller that feedback-controls process variables such as position, speed, and rotation speed.

【0002】

[Conventional technology]

A two-degree-of-freedom PID controller has been conventionally used in order to perform good feedback control of the control amount both when the disturbance is input and when the target value is changed.

【0003】

Here, the two-degree-of-freedom PID control is a PID control that simultaneously optimizes both the suppression characteristic for disturbance and the follow-up characteristic for a change in target value.

【0004】

FIG. 20 is a block configuration diagram of a conventional general two-degree-of-freedom PID or two-degree-of-freedom I-PD controller.

【0005】

This controller is composed of a 2-degree-of-freedom PID (hereinafter, including 2-degree-of-freedom I-PD) controller 100 and a controlled object 101, and this 2-degree-of-freedom PID degree controller 100 is derived from a proportional system, an integral system, and a differential system. It has become.

【0006】

In the proportional system, the target value R is multiplied by the two-degree-of-freedom parameter b for the proportional term of the coefficient element 111, the control amount Y is subtracted by the subtraction element 112 from this multiplication value, and the control amount Y is introduced into the addition element 113. On the other hand, in the integration system, the control deviation E between the target value R and the control amount Y is obtained in the comparison element 114, and the integration time T is obtained for the obtained control deviation E._IIntegrate with the integral element 115 with, and introduce it into the addition element 113. Further, the differential system multiplies the target value R by the two-degree-of-freedom parameter c for the differential term of the coefficient element 116 to derive the subtraction element 117, where the control amount Y is subtracted from the multiplication value, and the obtained subtraction value is obtained. Is differentiated by the differential element 118 having the differential time TD, and then introduced into the addition element 113.

【0007】

The subtraction element 112 output, the integration element 115 output, and the differentiation element 118 output thus obtained are added by the addition element 113, and then manipulated by multiplying this addition value by the proportional gain Kp of the proportional gain coefficient element 119. After obtaining the quantity MV, it is applied to the control target 101 together with the disturbance amount.

【0008】

In addition, as shown in Fig. 21, the feedforward control method has been adopted in which the target value change is detected, the operation amount change is judged in advance in order to cancel the influence, and the operation is executed in advance 2 A degree of freedom PID controller is used.

【0009】

This two-degree-of-freedom PID controller 100 inputs the target value R into the target value feedforward element 121, and uses the two-degree-of-freedom parameter b for the proportional term and the two-degree-of-freedom parameter c for the differential term. b + cT_Ds) Execute the operation, and transmit the obtained output. Further, after obtaining the control deviation E between the target value R and the control amount Y in the comparison element 114, the integration time T is obtained for this control deviation E._IIntegrate with the integral element 115 with, and introduce it into the subtraction element 122. On the other hand, in the differential system, it is proportional to the control amount Y, proportional to the differential element 123, and the differential operation formula (1 + T)._DDifferentiate using s) and introduce into the subtraction element 122. Here, the output of the proportional / differential element 123 is subtracted from the output of the integral element 115, and the proportional gain K of the proportional gain coefficient element 119 is added to this subtraction result._PIs multiplied, and this multiplication value and the output of the target value feedforward element 121 are added by the addition element 124 to obtain the operation amount MV to be applied to the control target 101 (not shown).

【0010】

Further, conventionally, as shown in FIG. 13, a two-degree-of-freedom PID controller to which a target value filter element 131 is added has been used.

【0011】

This two-degree-of-freedom PID controller 100 calculates the target value R by the following calculation formula using the two-degree-of-freedom parameters b and c of the target value filter element 131, and obtains the optimum target value tracking characteristic. Calculate the target value.

【0012】

(1 + bT_Is + cT_IT_Ds²) / (1 + T_Is) ...... (1) After the control deviation E is obtained by comparing the calculation target value and the control amount Y obtained as described above with the comparison element 114, the control deviation E is proportional to the control deviation E and the integral element 132 is {1 + 1 / (T)._Is)} is calculated. In addition, a differential operation is executed on the control amount Y with the differential element 133. Then, in the subtraction element 134, the output of the differential element 133 is subtracted from the output of the proportional / integrating element 132, and the obtained subtraction output is multiplied by the proportional gain Kp of the proportional gain coefficient element 119 to obtain the manipulated variable MV. It is applied to the controlled object 101 (not shown) together with the amount of disturbance.

【0013】

[Problems to be Solved by the Invention]

By the way, in all of the above-mentioned two-degree-of-freedom PID controllers, the PID parameter (proportional gain K) that optimizes the control response to disturbance._P, Integration time T_I, Derivative time T_DAfter adjusting), it is necessary to adjust the two degrees of freedom parameters b and c in order to optimize the follow-up characteristics to the target value change by trial and error. In addition, the control target model is constructed in advance, the simulation is executed under the evaluation function that is offline, and each parameter of the transfer function, that is, the PID parameter and the two-degree-of-freedom parameter b and c are designed by the optimum value search. Is being done.

【0014】

As a result, complicated adjustment work by a skilled person is required to obtain the PID parameter and the two-degree-of-freedom parameters b and c that provide the optimum control response to the disturbance and the target value change, and the adjustment work. There is a problem that requires a lot of time.

【0015】

The present invention has been made in view of the above circumstances, and is a two-degree-of-freedom control device that quickly and easily generates two-degree-of-freedom parameters for obtaining an optimum control response to a change in a target value without requiring skill. Is to provide.

【0016】

[Means for solving problems]

In order to solve the above problems, the present invention uses a control parameter and a two-degree-of-freedom parameter in a two-degree-of-freedom control device in which both disturbance suppression and target value followability are satisfactorily controlled with respect to a controlled object. The rise time σD related to the disturbance input control response, the rise time σR related to the target value change control response, and the integration time T which is the control parameter._IOr the integration time T_IAnd derivative time T_DIn this configuration, a parameter generation mechanism for generating the two-degree-of-freedom parameters that defines the target value change control response is provided.

【0017】

In addition to the above input items, the quadratic coefficient α2D of the disturbance input control response norm model and the quadratic coefficient α2R of the target value change control response norm model are used to set a value proportional to the value obtained by a predetermined calculation formula. , The target value change control response may be defined as a two-degree-of-freedom parameter.

【0018】

In the present invention, by taking the above means, predetermined input items, that is, the rise time σD related to the disturbance input control response, the rise time σR related to the target value change control response, and the integration time T which is the control parameter._IOr the integration time T_IAnd derivative time T_DThe quadratic coefficient of the disturbance input control response norm model is α2D and the quadratic coefficient of the target value change control response norm model is α2R. By executing the operation of, it is possible to generate a two-degree-of-freedom parameter that defines the target value change control response, and it is possible to omit the conventional trial-and-error adjustment work.

【0019】

In addition, another invention uses the transmission characteristic type of the controlled object, the dead time and the limit period, or the transmitted characteristic type of the controlled object and the standardized dead time to define the target value change control response. Is provided with a parameter generation mechanism for generating the above.

【0020】

By taking the above measures, the present invention can be graph-functioned in advance based on the transmission characteristic type, dead time and limit period of the controlled object, or the transmitted characteristic type and standardized dead time of the controlled object. The two-degree-of-freedom parameter can be generated quickly and easily, and has the same effect as that of the above invention.

【0021】

Further, in another invention, the attribute value of the first normative model and the attribute value of the second normative model that defines the target value followability in the disturbance suppression I-PD control system are input, and both of these attribute values are input. A function element that generates a coefficient value as a function, and an arithmetic element that generates a two-degree-of-freedom parameter that improves target value followability by using the coefficient value of this function element, the rise time and integration time of the first normative model. A parameter generation mechanism having the above is provided.

【0022】

The present invention generates a coefficient value as a function of the attribute value of the first norm model and the attribute value of the second norm model that defines the target value followability in the disturbance suppression I-PD control system, and this coefficient value is generated. , The rise time and integration time of the first norm model are used to generate 2 degrees of freedom parameters that improve target value followability, so 2 degrees of freedom are automatically adjusted without trial and error adjustment work. Integral parameters can be generated.

【0023】

BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, embodiments of the present invention will be described with reference to the drawings. (First Embodiment) Prior to explaining one embodiment of the two-degree-of-freedom control device according to the present invention, the basic concept for realizing the present device will be described with reference to FIG. (1) First, based on the input of the target value R and the control amount Y, the proportional gain K is the same as before._P, Integration time T_I, Derivative time T_DIn addition, the configuration is provided with a 2-degree-of-freedom PID controller 1 that calculates the manipulated variable MV by using the 2-degree-of-freedom parameter b for proportional terms and the 2-degree-of-freedom parameter c for differential terms. The 2-degree-of-freedom PID controller 1 corresponds to, for example, the 2-degree-of-freedom PID controller 100 in FIG. (2) Next, in the two-degree-of-freedom PID controller 1, the proportional gain KP, the integration time TI, and the differential time TD, which are PID parameters that optimize the control response to the disturbance, are adjusted in advance, and the two-degree-of-freedom PID Set to controller 1. (3) Furthermore, a new parameter generation mechanism 2 is provided, and two degrees of freedom parameters b and c are automatically calculated based on a certain input item A. (4) In this way, the two-degree-of-freedom parameters b and c calculated by the parameter generation mechanism 2 are generated and set in the two-degree-of-freedom PID controller 1.

【0024】

In order to concretely realize the above idea, it is important what input item A is used and what kind of operation is performed by the parameter generation mechanism 2 based on the input item A to obtain the parameters b and c. Come on.

【0025】

Therefore, the procedure for deriving the input item A and the calculation formula will be described below.

【0026】

Expressing the conventional general two-degree-of-freedom control system (see Fig. 20) with a transfer function, the configuration is as shown in Fig. 2 (a).

【0027】

In the figure, 1 is a 2-degree-of-freedom PID controller, 3 is a control target, 4 is an addition element, R is a target value, Y is a control amount, and D is a disturbance.

【0028】

That is, in the 2-degree-of-freedom PID controller 1, the target value R and the control amount Y are input, and the proportional gain K is a control parameter for the suppression characteristic against disturbance._P, Integration time T_I, Derivative time T_D, The manipulated variable MV can be obtained by the following arithmetic expression using the two-degree-of-freedom parameter b for the proportional term and the two-degree-of-freedom parameter c for the differential term.

【0029】

K_P{bR-Y + (1 / T_Is) (RY) + T_Ds (cR-Y)} ...... (2) The manipulated variable MV obtained by the above calculation formula is applied to the control target 3 together with the disturbance D.

【0030】

By the way, the two-degree-of-freedom PID control system shown in FIG. 2 (a) can be virtually represented by the control system shown in FIG. 2 (b) by performing equivalent block conversion. 5 is a feedforward control element, 6 is an addition / subtraction element, and 7 is a feedback control element.

【0031】

In this control system, the feedforward control element 5 is {K with respect to the input target value R._P(b + 1 / T_Is + cT_DThe transfer function s)} is calculated and the output W is transmitted. Further, the feedback control element 7 is {K with respect to the control amount Y which is the output of the control target 3._P(1 + 1 / T_Is + T_D)} Computes the transfer function and transmits the output V. Then, in the addition / subtraction element 6, the output V of the feedback control element 7 is subtracted from the output W of the feedforward control element 5, and the obtained subtraction output is applied to the control target 3 to generate the control amount Y. At this time, the disturbance D is additively added to the addition / subtraction element 6 and input to the control target 3.

【0032】

Also, the transfer function of control target 3 is G._PIf (s), the transfer function W from the disturbance D to the control quantity Y in the feedback control system_D(s) can be expressed by the following equation.

【0033】

W_D(s) = G_P(s) / [1 + G_P(s) K_P {1 + 1 / (T_Is) + T_Ds}] = (T_I/ K_P) s / [1 + {1 + 1 / (K_PG_P(s))} T_Is + T_IT_Ds²] ...... (3) On the other hand, the normative model that defines the desired control response of the control quantity Y by the disturbance D is as follows (PID control, p14 (2.11) or p31 (2.35), author: Nobuhide Suda. , Editor: System Control Information Society, Publisher: Asakura Shoten Co., Ltd.).

【0034】

M_D(s) = {s / (K_P/ T_I)} [1 / (1 + σDs + α2DσD²s² + α3DσD³s³)] ......(Four) Here, the coefficients α2D and α3D of the normative model for the control response to the disturbance input are set to the values shown in Table 1, for example, in consideration of the balance between the quick response and the robustness. For this Table 1, see the above-mentioned literature PID control, p14 to p15.

【0035】

[table 1]

<img file="JP2000214902A_D0001.tif" />【0036】

By the way, the actual control system of Eq. (3) and the normative model of Eq. (4) must have the same relationship. Therefore, W_D(s) = M_DSolving (s), the rise time σ regarding the control response to the disturbance input_DAnd PID parameter K_P, T_I, T_{D is}Can be decided. Therefore, if a step-like disturbance is applied to the input side of the controlled object to the PID feedback control system in which the determined parameters are set, a control response as shown in FIG. 7A can be obtained.

【0037】

Next, for the control system shown in FIG. 2 (b), the determined control parameter K_P, T_I, TD is set, and the transfer function W from the target value R at this time to the control quantity Y_RExpressed with (s), it becomes as follows.

【0038】

W_R(s) = [K_P{b + 1 / (T_Is) + cT_Ds}] M_D(s) ...... (5) Here, the inside of [] in the first stage is the transfer function of the feedforward control element 5, and the M in the second stage._D(s) is the transfer function of the normative model of Eq. (4) above.

【0039】

Therefore, this equation (5) can be expressed by the following equation (6).

【0040】

W_R(s) = (1 + bT_Is + cT_IT_Ds²) / (1 + σDs + α2DσD²s²+ α3DσD³s³) ...... (6) On the other hand, the normative model that defines the desired control response of the control quantity Y by the target value R shall be expressed by the following equation (see the above-mentioned document PID control, p120 (6.2) equation).

【0041】

M_R(s) = 1 / (σRs + α2RσR²s²+ α3RσR³s³) ... (7) Here, the coefficients α2R and α3R of the normative model for the control response to the change in the target value are set to the values shown in Table 1 above, for example, from the balance between quick response and robustness.

【0042】

So W_R(s) = M_RSolving (s), the rise time σR for the control response to the change in the target value can be obtained from the minimum real root of the cubic equation in the following equation.

【0043】

(1-2α2R + α3R) σR³-(1-α2R) σDσR² + α2DσD²σR-α3D σD³= 0 ...... (8) Here, W of equation (6)_RSet (s) = B / C, and M in Eq. (7)_RIf you set (s) = 1 / Z, W_R(s) = M_RFrom the relationship of (s) B / C = 1 / Z Z = C / B ...... (9) Can be derived.

【0044】

Therefore, if the C / B is calculated, the solutions are derived up to the first-order term and the second-order term, and the solutions are placed in the same relations with the first-order term and the second-order term of Z, respectively, the two-degree-of-freedom parameter b for the proportional term can be obtained. b = (σD-σR) / T_I ......(Ten) It can be determined from the arithmetic expression. The two-degree-of-freedom parameter c for the differential term is c = {(1-α2R) σR²-σDσR + α2DσD²} / T_IT_D = {-σR (σD-σR) + α2DσD²-α2RσR²} / T_IT_D ...... (11) It can be determined from the arithmetic expression.

【0045】

When α2D = α2R = 0.5 in Eq. (11), c = (σD-σR)²/ (2T_IT_D) ...... (12) And 2 degrees of freedom parameters b and c are (σ)_D-σ_R) Is a function.

【0046】

Therefore, as is clear from the above description, the input items A and the arithmetic expressions for obtaining the two degrees of freedom parameters b and c have been obtained, and specific examples thereof will be described below.

【0047】

FIG. 3 is a configuration diagram showing an embodiment of a two-degree-of-freedom PID controller according to the present invention.

【0048】

In the control device shown in FIG. 6A, the parameter generation mechanism 2 is connected to the 2-degree-of-freedom PID controller 1, and the parameter generation mechanism 2 has the integration time TI and the differentiation time T as input items A._D, The rise time σD that defines the disturbance input control response and the rise time σR that defines the target value change control response are input, and 2 degrees of freedom parameters b and c are calculated by the calculation means shown in Fig. 4 described later, and 2 degrees of freedom are calculated. The configuration is set to PID controller 1.

【0049】

In addition to the input items shown in Fig. (A), the control device shown in Fig. (B) also has the quadratic coefficient α2D of the disturbance input control response norm model and the quadratic coefficient α2R of the target value change control response norm model. The configuration is such that the additional input is performed, the two-degree-of-freedom parameters b and c are calculated by the calculation means shown in FIG. 5 described later, and the two-degree-of-freedom PID controller 1 is set.

【0050】

That is, the parameter generation mechanism 2 shown in FIG. 4A is composed of the comparison element 11 and the division element 12, and the comparison element 11 is the target value change control also input from the rise time σD that defines the input disturbance input control response. The rise time σR that defines the response is subtracted, and the subtracted output is introduced into the division element 12. This division element 12 divides the subtraction result (σD-σR) by the integration time TI, calculates the two-degree-of-freedom parameter b for the proportional term in the above equation (10), and sets it in the two-degree-of-freedom PID controller 1. ..

【0051】

The rise time σD that defines the disturbance input control response can be automatically determined when designing the PID parameter by the partial model matching method, or the step disturbance occurs when the PID parameter is determined using another method. Since the control amount Y is almost equal to the time when the peak value is reached at the time of entering, it can be determined from this time as well.

【0052】

Further, the rise time σR that defines the target value change control response can be uniquely determined by calculation as described above, or can be arbitrarily set as the time when the control amount at the time of changing the target value step reaches approximately 50%.

【0053】

Next, the parameter generation mechanism 2 shown in FIG. 4B is provided with a comparison element 13, a division element 14, a multiplication element 15, a division element 16, and a coefficient element 17. The comparison element 13 and the division element 14 calculate the two-degree-of-freedom parameter b for proportional terms in the same manner as in FIG. 4 (a). Further, in the multiplication element 15, the output of the comparison element 13 and the output of the division element 14 are multiplied, and the multiplication result is introduced into the division element 16, where the multiplication result is obtained with the derivative time T._DAfter dividing by, multiply by 1/2 by the coefficient element 17, and the two-degree-of-freedom parameter c = (σD-σR) for the differential term based on the above equation (12).²/ (2T_IT_D) Can be calculated and transmitted to PID controller 1 with 2 degrees of freedom.

【0054】

Further, FIG. 5 is a diagram showing the internal configuration of the parameter generation mechanism 2 of FIG. 3 (b).

【0055】

This parameter generation mechanism 2 is composed of a first arithmetic element 21 and a second arithmetic element 22, and the first arithmetic element 21 calculates the two-degree-of-freedom parameter b for proportional terms by the calculation of Eq. (10). 2 Degrees of freedom Call PID controller 1. The second arithmetic element 22 calculates the two-degree-of-freedom parameter c for the differential term by the operation of Eq. (11), and similarly transmits it to the two-degree-of-freedom PID controller 1.

【0056】

Next, another example of the parameter generation mechanism 2 will be described with reference to FIG.

【0057】

In Fig. 6 (a), the difference signal (σD-σR) of the rise time σD and σR that defines the response control by the disturbance input and the target value change is derived to the division element 12, and the difference signal is integrated here with the integration time T._IThis is an example of calculating the two-degree-of-freedom parameter b for proportional terms by dividing by.

【0058】

Fig. 6 (b) has two degrees of freedom with the same configuration as Fig. 4 (b), except that the difference signal (σD-σR) between the compatible uplink times σD and σR is input as in Fig. 4 (a). This is an example of calculating parameters b and c and transmitting them to PID controller 1 with 2 degrees of freedom.

【0059】

An example of the control response when the target value step is changed when the two degrees of freedom parameters b and c are set based on the input item A determined as described above is as shown in FIG. 7 (b). .. For comparison, the control response during conventional I-PD control (b = c = 0) is shown in Fig. 7 (c), and the control response during basic PID control (b = c = 1) is shown in Fig. 7 (c). As shown in d).

【0060】

As is clear from these control response diagrams, by setting b and c to arbitrary predetermined values without being limited to b = c = 0 or 1, without causing a large overshoot when the target value is changed. , 2 degrees of freedom PID control with excellent followability can be realized.

【0061】

In addition, σD, σR, T_I, T_DIs not limited to the values described above, and two degrees of freedom parameters b and c may be determined and set by the arithmetic expressions of Eqs. (10) and (11) using any determined value. ..

【0062】

Therefore, according to the above embodiment, if a predetermined input item is input to the parameter generation mechanism 2, the two degrees of freedom parameters b and c are set based on the above-described derived predetermined arithmetic expression. It can be set to the PID controller 1 with 2 degrees of freedom, and it is possible to save the trouble of determining the optimum parameters while adjusting by trial and error over a long period of time as in the past, speeding up the adjustment work and Parameters can be determined without relying on an expert.

【0063】

(Second embodiment) In the first embodiment, the input item A (σ) is based on the derivation process of equations (1) to (11)._D, Σ_{R, TI, TD,}σ2_D,σ2_R)Was selected, and the two-degree-of-freedom parameters b and c were determined by a predetermined arithmetic expression based on these input items A. However, in the present embodiment, the two-degree-of-freedom parameters b and c are set using a graph function. It is configured to be transmitted.

【0064】

That is, in the present embodiment, the parameters b and c obtained by the above-mentioned equations for various control targets having a lag order are shown in FIGS. 8 (a) and 8 (b). Therefore, a graph having a relationship as shown in FIG. 8 is set in advance in the parameter generation mechanism 2 as a function, and the transfer function (test batch) type of the controlled object, the waste time L, and the limit period input from the outside are set in advance. When a standardized waste time LN (= L / Tc) is received instead of Tc or L and Tc, the two-degree-of-freedom parameters b and c as a function may be transmitted.

【0065】

9 and 10 are block diagrams showing an embodiment of a two-degree-of-freedom PID controller according to the present invention.

【0066】

This control device is composed of a 2-degree-of-freedom PID controller 1 and a parameter generation mechanism 2 as in Fig. 3, but in Fig. 9, a graph function as shown in Fig. 8 is set in the parameter generation mechanism 2 and input. This is an example of generating two degrees of freedom parameters b and c based on the controlled object type to be controlled, the dead time L, and the limit period Tc.

【0067】

In Fig. 10, the graph function shown in Fig. 8 is set in the parameter generation mechanism 2, and the two degrees of freedom parameters b and c are also transmitted based on the input control target type and the standardized waste time LN (L / Tc). This is an example of setting to 2 degrees of freedom PID controller 1.

【0068】

Therefore, according to this embodiment, for various control targets, if the parameters b and c determined by the first embodiment are set in advance in the parameter generation mechanism 2 as graph functions, the control target type, By simply inputting input items such as wasted time, the optimum two-degree-of-freedom parameters b and c can be quickly generated according to the control target type, and the target value can be changed without the need for complicated adjustment work. On the other hand, a 2-degree-of-freedom PID controller with excellent followability can be realized. (Third Embodiment) This embodiment is an example applied to a two-degree-of-freedom I-PD controller instead of the two-degree-of-freedom PID controller. I-PD control is a control that operates depending only on the amount of control for both proportional operation and differential operation.

【0069】

Since this two-degree-of-freedom I-PD control system basically has the same configuration as in FIGS. 20 to 22, the two-degree-of-freedom I-PD control device has the same conceptual configuration as in FIG. It can be represented by a figure 11.

【0070】

That is, in the 2-degree-of-freedom I-PD controller, the parameter generation mechanism 2 is connected to the 2-degree-of-freedom I-PD controller 1', and this parameter generation mechanism 2 is for proportional terms and for proportional terms based on the input input item A. It is a mechanism that generates two degrees of freedom parameters b and c for differential terms.

【0071】

This two-degree-of-freedom PID control system can be represented by a block diagram as shown in FIG. 12 (a) from the conventional configuration of FIG. 20.

【0072】

In FIG. 12, 31 is a 2-degree-of-freedom I-PD controller (2-degree-of-freedom PID controller), 32 is a control target, 33 is an addition element, R is a control amount, Y is a control amount, and D is a disturbance.

【0073】

In this two-degree-of-freedom I-PD controller 31, the target value R and the control amount Y are input, and the proportional gain K is a PID parameter for the suppression characteristic against disturbance._P, Integration time T_I, Derivative time T_D, The manipulated variable MV is calculated by the following arithmetic expression using the two-degree-of-freedom parameter b for the proportional term and the two-degree-of-freedom parameter c for the differential term.

【0074】

K_P{bR-Y + (1 / T_Is) (RY) + T_Ds (cR-Y)} ......(13) The manipulated variable MV obtained in this way is applied to the control target 32 together with the disturbance D.

【0075】

By the way, the control system shown in FIG. 12A is virtually composed of the feedforward control element 34 and the feedback control element 35 as shown in FIG. 12B when the equivalent block is converted. 32 is the control target and 36 is the addition / subtraction element.

【0076】

In this control system, the feedforward control element 34 multiplies the input target value R by the transfer function of the control element 34 and transmits the output W. Further, the feedback control element 35 calculates the transfer function of the control element 35 with respect to the control amount Y which is the output of the control target 32, and transmits the output V. Then, the addition / subtraction element 36 subtracts the output V of the feedback control element 35 from the output W of the feedforward control element 34, applies the obtained subtraction output to the control target 32, and generates a control amount Y. At this time, the disturbance D is additively added to the addition / subtraction element 36 and input to the control target 32.

【0077】

Also, the transfer function of the controlled object 32 is G._PIf (s), the transfer function W from the disturbance D to the control quantity Y in the feedback control system_D(s) can be expressed by the following equation.

【0078】

W_D(s) = G_P(s) / [1 + G_P(s) K_P {1 + 1 / (T_Is) + T_Ds}] = (T_I/ K_P) s / [1 + {1 + 1 / (K_PG_P(s))} T_Is + T_IT_Ds²] ......(14) On the other hand, the normative model that defines the desired control response of the control quantity Y by the disturbance D is as follows (PID control, p14 (2.11) or p31 (2.35), author: Nobuhide Suda. , Editor: System Control Information Society, Publisher: Asakura Shoten Co., Ltd.).

【0079】

M_D(s) = {(T_I/ K_P) s} / (1 + σDs + α2DσD²s² + α3DσD³s³+ α4DσD⁴s⁴) ...... (15) Here, the coefficients α2D, α3D, and α4D of the normative model for the control response to the disturbance input are set to the values shown in Table 2, for example, from the balance between quick response and robustness.

【0080】

[Table 2]

<img file="JP2000214902A_D0002.tif" />【0081】

Here, W as above_D(s) = M_DSolved as (s), the rise time σD and PID parameter K for the control response to the disturbance input_P, T_I, T_DCan be determined. Assuming that a step-like disturbance is applied to the input side of the controlled object to the I-PD feedback control system in which the determined parameters are set, a control response as shown in FIG. 17 can be obtained.

【0082】

Next, for the control system shown in FIG. 12 (b), the determined control parameter K_P, T_I, T_DIs set, and the transfer function W from the target value R to the control amount Y at this time is set._RExpressed with (s), it becomes as follows.

【0083】

W_R(s) = [K_P{b + 1 / (T_Is) + cT_Ds}] MD (s) ...... (16) Here, the inside of [] in the first stage is the transfer function of the feedforward control element 34, and the MD (s) in the second stage is the transfer function of the normative model of the above equation (15).

【0084】

Therefore, this equation (16) can be expressed by the following equation (17).

【0085】

W_R(s) = (1 + bT_Is + cT_IT_Ds²) / (1 + σDs + α2DσD²s² + α3DσD³s³+ α4DσD⁴s⁴) ...... (17) On the other hand, the normative model that defines the desired control response of the control quantity Y by the target value R shall be expressed by the following equation (see the above-mentioned reference PID control, p120 (6.2) equation).

【0086】

M_R(s) = 1 / (1 + σRs + α2RσR²s²+ α3RσR³s³) ... (18) Here, the coefficients α2R and α3R of the normative model for the control response to the change in the target value are set to the values shown in Table 2 above, for example, in consideration of the balance between the quick response and the robustness.

【0087】

So W_R(s) = M_RWhen solved as (s), ignoring higher-order terms, the rise time ratio σR / σD between the rise time σR for the control response to the target value change and the rise time σD for the control response to the disturbance input is the cubic equation of the following equation. Obtained from the smallest real root.

【0088】

(1-2α_2R+ α_3R) (Σ_R/ σ_D)³-(1-α_2R) (Σ_R/ σ_D)² + α_2D(σ_R/ σ_D)-α_3D= 0 ...... (19) In the two-degree-of-freedom IP control that does not use the differential time, the rise time ratio σR / σD can be obtained as the minimum real root of the following quadratic equation.

【0089】

(1-α2R) (σR / σD)²-(ΣR / σD) + α2D = 0 ...... (20) Next, the coefficient β for calculating the two-degree-of-freedom parameter b for the proportional term is β = 1- (σR / σD) ...... (21) In addition, the coefficient γ for calculating the two-degree-of-freedom parameter c for the differential term is γ = (1-α2R) (σR / σD)²-(ΣR / σD) + α2D ...... (22) And.

【0090】

Therefore, using these values, the two-degree-of-freedom parameter b for proportional terms is b = βσD / T_I ......(twenty three) And the two-degree-of-freedom parameter c for the differential term is c = γσD²/ (T_IT_D) ......(twenty four) Can be decided.

【0091】

Therefore, a function element that generates these coefficients β and a coefficient γ is required, and the function element that generates these coefficients is created as follows.

【0092】

That is, using the normative model MD (s) of equation (15) determined from a certain interpolation coefficient λD and the normative model MR (s) of equation (18) determined from a certain interpolation coefficient λR, the above equation (19) The rise time ratio σR / σD is obtained from the equation or the equation (20), and the coefficients β and γ are obtained from the equations (21) and (22).

【0093】

Then, the coefficients β and γ are obtained for various values of the interpolation coefficients λD and λR, and a matrix data table in which the x-axis is λD, the y-axis is λR and the z-axis is β and γ is created, and the interpolation coefficient λD is created. By inputting, λR, a function element that generates coefficients β and γ can be created.

【0094】

Since there is a relationship shown in FIG. 15 between the interpolation coefficient λ and the maximum sensitivity Ms, the maximum sensitivity MSD and MSR may be used instead of λD and λR. When a quadratic norm model is used, the coefficient α2R of the denominator quadratic term of the norm model may be used instead of λR.

【0095】

Further, as an example of the function element, for example, as shown in FIG. 16A, if the x-axis is λD of the cubic norm model and the y-axis is α2R of the quadratic norm model, the coefficient β for 2-degree-of-freedom IP control is set. You will get a function to find. As shown in FIGS. 16 (b) and 16 (c), if the x-axis is λD of the quaternary norm model and the y-axis is λR of the cubic model, the coefficient β for 2-degree-of-freedom I-PD control , A function for finding γ is obtained.

【0096】

FIG. 13 is a configuration diagram showing an example of a parameter generation mechanism embodied based on the above explanation.

【0097】

This parameter generation mechanism 2 is composed of a function element 41 and an arithmetic element 42. This function element 41 includes the attribute value XD (λD, MSD, etc.) of the normative model of the I-PD control system for disturbance suppression and the attribute value XR (γR, MSR, α2R, etc.) of the normative model that defines the target value followability. It is a matrix data table in which the values of the coefficient β are stored for various sets of values of, and the coefficient β stored in the matrix data table is generated based on the input attribute value XD and the attribute value XR. Instead of the matrix data table, it may be an element that generates the coefficient β by the calculation of the above equations (19) or (20) and (21).

【0098】

The arithmetic element 42 is composed of a multiplication element 43 and a division element 44, of which the multiplication element 43 is divided by multiplying the coefficient β generated from the function element 41 by the rise time σD of the normative model of the disturbance suppression I-PD control system. Introduce to element 44. This division element 44 integrates the multiplication value obtained by the multiplication element 43 into the integration time T._IDivide by and calculate the two-degree-of-freedom parameter b for proportional terms, and set it in the two-degree-of-freedom I-PD controllers 1'and 31.

【0099】

FIG. 14 is a configuration diagram showing another example of the two-degree-of-freedom parameter generation mechanism.

【0100】

This parameter generation mechanism 2 is composed of two function elements 45 and 46 and two arithmetic elements 47 and 48 corresponding to these function elements 45 and 46, respectively.

【0101】

These function elements 45 and 46 are matrix data tables in which the values of the coefficients β are stored for various sets of values of the attribute values XD and XR of the normative model, respectively, and the input attribute values XD and the attribute values are stored. Based on XR, the coefficients β and γ stored from the matrix data table are generated, respectively. Instead of the matrix data table which is the function element 45, the coefficient β is generated by the operation of the above equation (19) or the equations (20) and (21), and instead of the matrix data table which is the function element 46. In addition, it may be an element that generates a coefficient γ by the calculation of the above equations (19) and (22).

【0102】

The arithmetic element 47 is composed of a multiplication element 49 and a division element 50 like the arithmetic element 42 shown in FIG. 13, and the multiplication element 49 has a coefficient β generated from the function element 45 as a normative model of the disturbance suppression I-PD control system. Multiplies the rise time σD of and introduces it to the division element 50. The division element 50 calculates the two-degree-of-freedom parameter b for the proportional term by dividing the multiplication value obtained by the multiplication element 49 by the integration time TI.

【0103】

One arithmetic element 48 is composed of a first multiplication element 51, a second multiplication element 52, a division element 53, and a multiplication element 54, and the two-degree-of-freedom parameter c for the differential term is calculated by the calculation of the above equation (24). To do. Specifically, rising obtained in the second multiplier element 52 multiplies values of Ri time .sigma.D .sigma.D²And the coefficient γ are multiplied by the multiplication element 51 and introduced into the division element 53. This division element 53 is the integration time T obtained by the multiplication element 54 by multiplying the multiplication value from the multiplication element 51._IAnd derivative time T_DDivide by the multiplication value with, and calculate the two-degree-of-freedom parameter c for the differential term by Eq. (24). Then, the parameters b and c obtained by the arithmetic elements 47 and 48 are transmitted to, for example, the two-degree-of-freedom I-PD controller 1'shown in FIG.

【0104】

The two-degree-of-freedom I-PD controller 1'is not limited to the one shown in FIG. 20, but is the target value feedforward type shown in FIG. 21 or the target shown in FIG. 22 obtained by converting this into an equivalent block. It may be a value filter type, a loop compensation type, a feedback compensation type, etc. (PID control, p75, author: Nobuhide Suda, Asakura Shoten).

【0105】

Further, although the differential element 118 shown in FIG. 20 is described as a complete differential type, it may be an inexact differential type to which a first-order lag element is added.

【0106】

Next, the input to the parameter generation mechanism will be described. When the PID value is designed using the norm model so that the disturbance can be suppressed best, the interpolation coefficient λD, the maximum sensitivity MSD, the coefficient α2R of the denominator quadratic term, and the rise time σD are unique in the norm model used at this time. Since it is decided, use these. When the PID value is adjusted without using the normative model, the interpolation coefficient of the waveform with the closest shape is compared with the control response (Fig. 17) drawn by changing the interpolation coefficient λD when step input is applied to the operation end. From λD, identify and use the normative model and rise time σD (time when the control response almost peaks).

【0107】

Note that FIG. 17 (a) is a control response diagram of the fourth-order normative model, which is used when a two-degree-of-freedom I-PD control system using differential motion is used. In addition, Fig. 17 (b) is a control response diagram of a cubic norm model, which is used for a two-degree-of-freedom IP control system that does not use differential motion.

【0108】

Also, if you want to reduce the change in the amount of operation that occurs when the target value is changed excessively, or if you want to increase the robustness, set the rise time σD to a value larger than the value determined by the above, and if you want to speed up the response, set it to a smaller value. Enter the value.

【0109】

In addition, for example, a mechanism for automatically identifying the normative model of the I-PD control system has been added, and the integration time T used in the I-PD controller has been added._I, Derivative time T_DIf the input items of the identification result are automatically input to the parameter generation mechanism of the present invention, an auto (self) tuning 2 degree of freedom I-PD control system can be configured.

【0110】

Furthermore, as a normative model that defines the followability of the target value change, refer to the control response waveform (Fig. 18) drawn by changing the interpolation coefficient λR and the coefficient α2R of the denominator quadratic term, and from the overshoot amount, etc. Determine and input the interpolation coefficient λR that gives the most appropriate shape in the control system and the coefficient α2R of the denominator quadratic term, that is, the attribute value of the normative model. Note that FIG. 18 (a) is a control response diagram of the cubic norm model when the interpolation coefficient λR at the time of changing the target value unit step is variously changed, and is used for 2-degree-of-freedom I-PD control. b) is a control response diagram of the quadratic norm model when the coefficients of the denominator quadratic term when the target value unit step is changed are changed, and is used when IP control with two degrees of freedom is performed.

【0111】

FIG. 19 is an example of a control response when a parameter generation mechanism 2 according to the present invention generates a 2-degree-of-freedom parameter, sets it in a 2-degree-of-freedom IP (D) controller, and controls a controlled object. For comparison, the control response is shown for the case of control by one degree of freedom IP (D) controller and PI (D) controller.

【0112】

As is clear from this figure, in the case of the I-PD control according to the present invention, it is possible to respond with quick response without generating a large overshoot when the target value is changed, and the target value followability is greatly improved. Can be improved.

【0113】

[Effect of the invention]

As described above, according to the present invention, two degrees of freedom parameters that specify the control response when the target value is changed can be automatically generated based on a predetermined input item, and a trial takes a long time as in the conventional case. The work of adjusting by mistake can be omitted, and the two-degree-of-freedom parameter can be generated in a short time without relying on an expert.

【0114】

Therefore, it is possible to provide a two-degree-of-freedom control device that can quickly realize optimum controllability against both disturbance and change of target value.

[Simple explanation of drawings]

[Figure 1]

The conceptual block diagram explaining the 1st and 2nd Embodiment of the 2 degree of freedom control apparatus which concerns on this invention.

[Figure 2]

The block diagram of the two-degree-of-freedom PID control system explaining the principle of this invention.

[Fig. 3]

The whole block block diagram including the input item in 2 degrees of freedom control device which concerns on this invention.

[Fig. 4]

The figure which shows the concrete configuration example of the parameter generation mechanism shown in FIG.

[Fig. 5]

The figure which shows the concrete configuration example of the parameter generation mechanism shown in FIG.

[Fig. 6]

The figure which shows the concrete configuration example of the parameter generation mechanism shown in FIG.

[Fig. 7]

The figure which shows the example of the control response when the 2 degree of freedom control device which concerns on this invention is used.

[Fig. 8]

The figure explaining the function set in the parameter generation mechanism shown in FIG.

[Fig. 9]

The block block diagram which shows the 2nd Embodiment of the 2 degree of freedom control apparatus which concerns on this invention.

[Fig. 10]

The block block diagram which shows another example of 2nd Embodiment of the 2 degree of freedom control apparatus which concerns on this invention.

[Fig. 11]

The conceptual block diagram explaining the 3rd Embodiment of the 2 degree of freedom control apparatus which concerns on this invention.

[Fig. 12]

A block diagram of a two-degree-of-freedom I-PD control system illustrating the principle of the present invention.

[Fig. 13]

The figure which shows the concrete configuration example of the parameter generation mechanism.

[Fig. 14]

The figure which shows the concrete configuration example of the parameter generation mechanism.

[Fig. 15]

The figure which shows the relationship between the interpolation coefficient of a normative model and the maximum sensitivity.

[Fig. 16]

The figure which shows the function characteristic of a function element.

[Fig. 17]

The figure which shows the control response of the normative model at the time of a step disturbance input used for PID value design or characteristic identification of an I-PD control system.

[Fig. 18]

2 Degree of freedom A diagram showing the control response of the norm model when the unit step is changed for the target value used in the control design.

[Fig. 19]

The figure which shows the example of the control response when the 2 degree of freedom control device which concerns on this invention is used.

[Fig. 20]

The figure which shows the schematic structure of the commonly known 2 degree of freedom PID or 2 degree of freedom I-PD control system.

[Fig. 21]

A block diagram of a two-degree-of-freedom control device having a conventionally known feedforward element.

[Fig. 22]

A block diagram of a two-degree-of-freedom control device having a conventionally known target value filter element.

[Explanation of symbols]

1 ... 2 degrees of freedom PID controller 1'... 2 degrees of freedom I-PD controller 2 ... Parameter generation mechanism 21, 22 ... Arithmetic elements σD ... Rise time for disturbance input control response σR ... Rise time for target value change control response T_I... integration time T_D... differential time α2D ... quadratic coefficient of disturbance input control response norm model α2R ... Target value change control Response norm model quadratic coefficient 41,45,46 ... Function elements 42,47,48 ... Arithmetic elements