JPS62114004A - Feedback control system - Google Patents

Abstract

PURPOSE:To execute simultaneously an optimum control to a target value follow-up and a disturbance suppression by providing a PID type transfer function arithmetic part and a PD type transfer function arithmetic part, and using the sum of output results of these two arithmetic parts, as a control output. CONSTITUTION:To a PID type transfer function arithmetic part 15, a deviation of a target value given from a target value setting part 13, and a process variable value given through a detecting element 12 of a process 2 is led, and on the other hand, an optimum PID parameter for a target follow-up is given from a control parameter setting part 14, and a transfer function G1(S) is calculated. On the other hand, a PD type transfer function arithmetic part 16 executes an operation of a transfer function G2(S), based on the process variable value given through the detecting element 12, and an optimum PID parameter for a disturbance suppression from the control parameter setting part 14. Results of operation of both the arithmetic parts 15, 16 are added by an adding part 17, and become a control signal to a final control element 11.

Description

[Detailed description of the invention] [Industrial application field] This invention uses P (proportional), which is the basis of continuous control.

(2) This invention relates to an improvement of the so-called PID control system, which performs control by combining various calculations of (integral) and D (differential).

[Conventional technology]

FIG. 3 is a block diagram showing a general PID control system.

As is well known, P
, I and D all kinds of operations now P I D
filiJIfJ&tlvintermediate, Teflo*x 2(D
This method was introduced in consideration of target value tracking and disturbance suppression.

There have been many research presentations on optimal adjustment methods for PID control, but the common conclusion of all of them is that the optimal PID parameters for "target value tracking" and "disturbance suppression" are different from each other. .

It has been pointed out that

However, since the current PIDli# meter can only set one type of control constant (parameter), it is not possible to simultaneously satisfy both the requirements of "target value tracking" and "disturbance suppression". In other words, the PID of "target value tracking"
This is because if a parameter is used, settling against a disturbance will be delayed, and conversely, if a PID parameter for "disturbance suppression" is used, an overshoot will occur when following a change in the target value.

For this reason, an I-FD control system as shown in FIG. 4 has also been proposed.

I-PD sub-control was designed exclusively to eliminate overshoot during target value tracking, and the PID control method performs all types of calculations of P, I, and D regarding deviation (r-x). As shown in the figure, the deviation (r-x
) feedback is ■ operation (■ operation section 4 reference)
The PD operation (see PD calculation section 3) is performed using the variable
Do it with chisel. This I-PD sub-control optimum control parameter is also determined, and the same parameter is used for "target value tracking" and "disturbance suppression".

[Problem that the invention seeks to solve]

However, the I-PD control method has a problem in that the responsiveness of the "target value tracking" control is significantly slower than in the case of PID control.

Therefore, this invention provides "target value tracking" and "disturbance suppression".
The purpose of this invention is to provide a control method that can simultaneously achieve both objectives by combining the optimal PID parameters.

[Means for solving problems]

A PID type transfer function calculation unit that receives the deviation between the target value and the process variable value as input, and a PD type transfer function calculation unit that receives the process variable value as input are provided, and the sum of the output results of these two calculation units is controlled. Control is performed as output.

[Effect]

The general formula for PID control is expressed by the following formula (1).

・・・・・・(1) Ko-K.

X: Process variable value (measured value) r: target value U: Manipulated variable value (control output) and KI-K(/TI KD-KCTD).

Also, the subscript r (Kcre Kxr * KDr ) f is added to the optimal PID parameter for "target value tracking", and the subscript d (Kcd - KId + KDd ) "it" is added to the optimal PID parameter for "disturbance suppression". I will represent each of them by attaching them.

By the way, when parameters for disturbance suppression are generally used, overshoot occurs during target value tracking because there is the following magnitude relationship between the two parameters.

Kc r < Kcd KDr < KDd This invention focuses on this point, and as shown in FIG.
The transfer functions Gl(S) and G2(8) of each part 1 and 3 are determined as shown in the following equation. Note that C is a constant that is sufficiently smaller than 1.

Gl(S) ・・・・・・(2) Gz(S)=(Kca(1-g)Kcr)+(K
nd-(1-II)Knr) S・・・・・・(3
) Furthermore, from the correspondence between equation (2) and equation (1), the relationship between each parameter is Kc-(1-#)Kcr'rD=KDr/Kcr=TDrTI-(
1l)Kcr/KIaM=(1-LiTld°
A relationship such as Kc r /Kcd holds true.

Therefore, by determining the transfer function as shown in equations (2) and (3), #! The control output u (S) synthesized with the polarity shown in Figure 1 is: u (8) = G1 (8) (r-x) G2 (8) x = Gt (S) r CGI (8) + G2 (8 )'l In other words, during disturbance suppression, since r is constant, if we ignore the Gt(S)rO term, (Gl(
Only the term 8)+02(S))X will work, but the part in the parentheses of this term is
2) and (3), which is exactly the PID calculation formula for disturbance suppression.

On the other hand, when tracking the target value, the integral parameter of equation (2) is set to K
Since Kld is used instead of lr, a difference will occur, but even if this difference becomes a problem, by adjusting the value of e, the The response can be similar. child\
One possible method for determining the value of C is to obtain the control output as shown in the following equation so that the control output until the process starts responding when changing the target value is the same as in the ideal case.

That is, (1g)Kcr+KId-L+2(1-1)KDr-K
cr+KIr-L+2KDr-'-g-(KId KIr)L/Kcr+2Kn
It can be obtained as r. Here, L represents the dead time t of the process. Note that although ξ is a value close to zero, even if it is actually set to -0, the control performance can be considerably improved compared to the conventional one.

[Implementation row]

FIG. 2 is a block diagram showing a first embodiment of the present invention. In the figure, 2 is a process, 11 is an expansion operation end, 12 is a detection end for detecting process variable values (measured values), 13 is a target value setting section, 14 is a control parameter setting section, and 15 is a PID type transfer function calculation section. , 16 is a PD type transfer function calculation section, and 17 is an addition section.

That is, the deviation between the target value given from the target value setting part 13 and the process variable value given via the detection end 12 of the process 2 is derived to the calculation part 15, while the deviation for target tracking is derived from the control parameter setting part 14. Since the optimal PID parameter of is given, the transfer function shown by the above equation (2) is calculated here.

On the other hand, the calculation section 16 calculates a transfer function such as the above equation (3) based on the process variable value passed through the detection end 12 and the optimum PID parameter for disturbance suppression from the control parameter setting section 14. The calculation results of both calculation units 15 and 16 are added by an addition unit 17, and become a control signal for the operating end 11.

Note that the transfer functions shown in equations (2) and (3) above can be changed as shown in (a) or (b) below.

(a) G1(S) = Kcr+Kxr/S+Knr 8G2
(S)-β(Kca Kcr) 10β(KpdKDr
)S (constant close to β-1) Gl (S)-KC+KIr /S+KDBG2(S)
-KC+KD'S indicates the control parameters. ) [Effects of the Invention] As described above, according to the present invention, optimal control for both target value tracking and disturbance suppression can be achieved simultaneously with one control device, resulting in a significant improvement in feedback control. This brings about the advantages that are possible.

[Brief explanation of drawings]

Fig. 1 is a main diagram that clearly expresses the feature f:M of this invention, Fig. 2 is a block diagram showing an embodiment of this invention, Fig. 3 is a block diagram showing a general PID control system, and Fig. 4 is a block diagram showing an embodiment of the invention. FIG. 1 is a block diagram showing a general I-PD control system. Symbol explanation: 1... PIDflllmltt, 2. Song/process, 3. Song PD calculation section, 4... ■ calculation section, 11.. Song operation end, 12..... Detection end, 13・
...Target value setting section, 14... Control parameter setting section, 15.16... Transfer function calculation section,
17... Addition section. Agent Patent Attorney Akio Namiki Agent Patent Attorney Kiyoshi Matsuzaki Figure 1 Figure 2

Claims (1)

[Claims] A feedback control method that performs feedback control of process values, comprising: a PID type transfer function calculation unit that receives as input the deviation between a target value and a process variable value, and a PD type that receives as input the process variable value. a transfer function calculation unit, the sum of the output results of each calculation unit is used as a control output, and the parameters of each transfer function are determined from each optimum PID parameter for target value tracking and disturbance suppression. Features a feedback control method.