JPH03214202A - Controller - Google Patents

Abstract

PURPOSE:To prevent the overshoot of the process value by obtaining the auxiliary set value via an auxiliary control part and correcting the set value based on the auxuliary set value. CONSTITUTION:A changing rate control part 10 outputs the set value SP serving as a single which varies to the target set value TSP from zero within a prescribed period of time after input of the value TSP. An auxiliary control part 11 outputs the auxiliary set value SSP and a selection signal SEL after receiving the input of the value TSP and the value SP. Then a selection part 12 selects the inputted value SP or SSP based on the signal SEL and outputs the selected value to a PID control arithmetic part 1 as the set value. Thus the overshoot of the process value can be automatically prevented.

Description

DETAILED DESCRIPTION OF THE INVENTION <Industrial Application Field> The present invention effectively suppresses process amount overshoot in a controller that controls an electric furnace or the like.

<Conventional technology> FIG. 12 shows the configuration of the controller.

TSP, which is a target set value for the process amount, and the process amount are input to the PID control calculation unit 1. PID control calculation section 1
performs proportional, integral, and differential calculations on the deviation between the target set value TSP and the process amount, calculates the manipulated amount, and outputs the calculated amount to the process 2, for example, an electric furnace. The process amount PV, which is the output of the process 2, is input to the prDof control calculation unit 1. In this way, the process amount is finally made to match the target set value "rsp."Incidentally, the target set value 'I'SP is manually increased (or decreased) at a constant rate by the change rate regulating section. There is also a configuration in which the setting value i is obtained and this setting value is input to the PID control calculation section 1.

In such a controller, overshoe of process quantity 1-
It is important to prevent In other words, the process quantity asymptotically approaches the target setting value 'I''SP, and this 'r' sp
Figure 13<A)
~(D) represents an example of control, P■ represents the process amount, OUT represents the change in the manipulated amount, and time t. It is assumed that the target set value TSP is applied at

Immediately after the target set value T”S P is applied, operate 10.
UT is output, and the process quantity PV asymptotically approaches the target set value ``rsp.'' By gaining experience, the operator can judge the presence or absence of overshoot h from the state of change in the process quantity P■. For example, in the case of (A), it is possible to determine the risk of overshoot (represented by NG), and in the case of (B), it is possible to determine the risk of overshoot (represented by G). ,(A>,
Even at times (C) and (D) different from (B), (C)
It can be determined that there is a risk of overshoot (NG), and that CD> should not occur (G). FIG. 14 shows an example in which the response is faster than that in FIG. 13. In this case as well, the operator can judge from experience whether (A) and (C> are at risk of overshooting (NG) and CB> and (D) should not be done (G). When a danger is detected, take measures such as lowering the voltage of the electric furnace heater power supply to prevent overshoot.

<Problems to be Solved by the Invention> However, in such a controller, the operator is able to detect the risk of overshooting the process amount through experience, so the operator must be on hand at all times, which can save labor. There was an issue that there was no such thing.

Additionally, since experience is required to detect the danger of overshoot, it is difficult for beginners to use.

<Object of the Invention> An object of the invention is to provide a controller that can prevent overshoot of the process amount.

In order to solve the above problems, the present invention calculates an auxiliary setting value by monitoring the setting value of the process amount to the target setting value or the setting value by the auxiliary control unit, and calculates the auxiliary setting value. By outputting the value to the control calculation unit, overshoot of the process amount is prevented.

Function> Prevents process amount overshoot by monitoring the settling status of the process amount and correcting the set value.

<Example> FIG. 1 shows an embodiment of a controller according to the present invention.

Note that the same elements as in FIG. 12 are denoted by the same reference numerals, and explanations thereof will be omitted. In FIG. 1, reference numeral 10 denotes a rate-of-change regulating section, which receives an mth set value TSP and outputs a set value SP, which is a signal that changes from zero to a target set value TSP in a predetermined period of time. 11 is an auxiliary control section, which controls the target set value TSP.
, the set value SP and the process amount are input, and the auxiliary set value S
Outputs SP and selection signal SEL. Reference numeral 12 denotes a selection section, into which the set value SP and the auxiliary set value SSP are input, selects one of them based on the selection signal SEL, and outputs it to the PID control calculation section 1 as a set value.

Next, the operation of this embodiment will be explained. The auxiliary control unit 11 determines whether an overshoot of the process amount is likely to occur (NG) or not (G) based on the following equations (1) and (2).

When DV≧-DPV, G ・・・・・・・・・(1
) NG when DV<K −DPV ・・・・・・(
2) DV2 Deviation between target set value T'SP (or set value SP) and process amount PV DPV: Change in process amount PV over a given time: Constant Figure 2 shows an example of control. Yes, PV is the process variable, OUT is the manipulated variable, TSP is the target set value, L,
A G represents the equivalent waste time of the process, which is the time between when the manipulated variable is input to the process and when the process variable changes.If =2, then in the case of (A), DV<K at the tip. -D PV , the judgment is NG based on the above equation (2), and 5 overshoots may occur. In (B), DV>K −D PV at the tip, and from equation (1) above, the determination is G, and it is determined that overshoot does not occur. Note that the predetermined time t1
As, 1/2 of the process equivalent dead time LAG was used. The determinations in equations (1) and (2) above were derived based on empirical rules. When the process amount P■ has not risen sufficiently, that is, when DV is large, the judgment is always G.
Therefore, there is no risk of overshoot. If the determination is NG, the auxiliary control unit 11 converts the set value SP to the target set value T.
The value corrected to move away from S P is output as the auxiliary set value SSP, and when determining G, the target set value 'I' S
Output a value corrected to be closer to F).

It is obvious that this operation can prevent overshoot of the process amount PV. Note that at this time, the selection signal S
EL is sent to cellar 1 so as to always select the auxiliary set value S S P side. Therefore, the selection section 12 and the selection signal S E
1. , is not a required element.

Figure 3 is an enlarged view of the state at the tip of the control, where the set value SP is at a constant slope and the target set value 'T' S P
The process amount PV is about to follow this set value SP. As shown in the figure, DV is K, DPV (
K=2), the determination is G based on the above equation (1). In this case, the selection signal SEL is set so that the selection unit 12 selects the set value SP. Auxiliary setting value SSP
should be between the target set value TSP and the set value SP.If the 0 judgment is G, the auxiliary set value SSP is made to match the set value SP, thereby eliminating the need for the selection unit 12 and the selection signal SEL.

The following equation (3) can also be used as the value of the auxiliary setting (tls sp ).

5SP= ('I'5P-K -DPV) ・p
+.

(3) K2: Constant (for example, 2) P: Constant for bipolar adjustment, -1 or 1 In this case, the selection unit 12 selects between the set value SP and the auxiliary set value SSP. ,
The value that is farther away from the target set value "I' S P is selected. After the set value SP approaches the target set value TSP and exceeds the auxiliary set value SSP, SSP is set even if DPV is constant. If -DPV becomes a constant value, it will naturally approach the target set value SSP according to the above equation (3).In this case, the auxiliary setting section 11 sets the auxiliary set value SSP so as not to satisfy the above condition. If this is set, the selection unit 12 and the selection signal SEL become unnecessary.Furthermore, if the rate of change regulation unit 10 is not provided, that is, in the case of 5P-TSP, the auxiliary set value SSP is selected from the beginning.Control in this case The situation is shown in Fig. 4(A).The auxiliary set point SSP first decreases and then approaches the target set value 'l''SP.
In this case, it is always determined to be G, and opacity xl~ does not occur.

However, as shown in CB), if a state is determined to be NG midway through, it is necessary to switch the auxiliary setting value SSP by the selection unit 12 to change it to the state of G.

In the above embodiment, the correlation between DV and DPV was judged in two states, G and NG, but the auxiliary setting value SSP
When using it for operation, it is better to evaluate the state in more detail.For that purpose, use fuzzy # reasoning to derive the auxiliary setting value SSP (for example, process amount F)V? DPV when exceeding auxiliary setting fNSsp
, DV, DSP (-TSP-8P) are set as 100%, SSP is corrected, and the target setting value is finally determined.
i''sp.For this purpose, an inference rule as exemplified below is used.

f DSP is PHand DV is 10a
nd DPV is N5thenΔSSP is P
H−・−・−−−−−−(4) f DSP +s Pa
and DV is PS and DPVis 7
0thenΔSSP +s Pa −−−−−−・−
-----(5) The meaning of rule (4) is 1-If, DS
If P is positive and medium, Dν is zero, and DPV is negative and small, ΔSSP is positive and medium. '', and rule (5) is ``If DSP is positive and large, and Dν is positive and small,
And if DPV is 17, ΔSSP is made positive and large.

” means something. An example of a combination of inference rules is shown in FIG. 5, and a membership function of the antecedent part is shown in FIG. 6. Using the combination of these rules and the membership function, a correction value ΔSSP of the auxiliary setting value SSP is determined. FIG. 7 shows a fuzzy inference method using the membership function of the consequent and the height method. The rules are as described above (4)
), according to equation (5). Note that the values of DSP, DV, and DPV are 70%, 5%, and -10%, respectively. ■ is mentioned above (
4) This is based on the rules of Eq. DSP is O16 due to membership function PM, and DV is 0.7 due to ZO.
5. DPV is calculated as 0.5 by the same NS, and the minimum value of these 0.5 is obtained.6ΔS S P is PM, so it is 5
It becomes 0%. ■ is based on the rule of equation (5) above, and DSP is 0. l, D
V is 0.25 with the same PS, and DPV is 0.25 with the same ZO.
5 is obtained. Since ΔSSP is PB, it becomes 100%.

The final correction value is determined by proportionally distributing the ΔSSP determined by these two rules using the following formula.

ΔS S P = (50% x O, 5+-100% x
o, 25) / (0,5 + 0.25) = 66.7%
In this way, the correction value Δ of the auxiliary setting (dis p
It is possible to obtain the value of SSP of 66.7%.

Figure 8 shows an example of control, where PV is the process amount and OUT
is the manipulated variable, and SSP is the auxiliary setting value.

As shown in (A), SSP initially decreases and then gradually approaches the target set value TSP. Therefore, the process amount P■
converges to the target setpoint TSP without overshooting. (B) shows the operation when a disturbance occurs. When the process amount P decreases due to disturbance, the auxiliary set value SSP
also decreases, and then gradually approaches the target set value 'Fsp. Therefore, the process amount P■ will not overshoot. FIG. 9 is a characteristic diagram comparing the control according to the present invention and the conventional control. PV-1 represents the change in the process amount due to the conventional control, and PV-2 represents the change in the process amount due to the control of the present invention. In the conventional control, when the set value SP becomes a constant value (A>, (B) and when a disturbance occurs (C
) both overshoot occurs, but according to the control of the present invention, overshoot does not occur. Fig. 10 is an example of another step response, and (A, ) is the response waveform of the process amount PV according to the present invention. . No overshoot occurs. (B) is an example of a conventional response waveform and is shown for comparison. An overflow occurs in the process amount PV.

FIG. 11 shows another embodiment of the present invention. Note that the same elements as in FIG. 1 are given the same reference numerals and their explanations will be omitted. 11th
In the figure, reference numeral 13 denotes a tuner, into which the process amount P■ and the tuning command are input, and outputs parameter data and output commands to the PID control calculation section 1 and auxiliary parameter data to the auxiliary control section 11. The tuner 13 selects the PI for use in the PIDID control section 3 section 1 according to the characteristics of the process 2.
This is for tuning the D parameter. When a tuning command is input to the tuner 13, the tuner 13
outputs an output command to the PID control calculation section 1. PID#
The Jm performance section 1 applies a predetermined disturbance to the process 1 according to this command. The tuner 13 adjusts the process amount P due to this disturbance.
Analyze the response in (2), calculate control parameters and auxiliary parameters, and output them to the PID control calculation unit 1 and auxiliary control unit 11, respectively. The control parameters include PID control constants, and the auxiliary parameters include equivalent dead time and time constant. The auxiliary control unit 11 calculates the selection signal SEL and the auxiliary setting value SSP from these auxiliary parameters.

By doing so, the configuration of the auxiliary control section 11 can be simplified and the auxiliary parameters can be automatically set.

In addition, since the characteristics of the process and the PID parameters are set based on known relationships, the integral time value is used instead of the equivalent dead time and time constant in order to calculate the predetermined time 'F1 in equations (1) and (2) above. Or you can use the differential time value,
It is possible to obtain these time values from the PID control calculation section 1.

<Effects of the Invention> As described above in detail based on the embodiments, in this invention, an auxiliary set value is obtained by the auxiliary control unit, and the set value is corrected using the auxiliary set value, thereby preventing overshoot of the process amount. It is designed to prevent this.

Therefore, what was previously controlled by the operator by determining the overshoot of the process amount, can now automatically predict and prevent overshoot, resulting in labor savings. There is.

It also has the effect of making it possible to control it without requiring any skill.

[Brief explanation of drawings]

FIG. 1 is a configuration diagram, not showing an embodiment of a controller according to the present invention;
Figures 2 and 4 are characteristic curve diagrams for explaining its operation, Figures 5 to 7 are diagrams for explaining fuzzy inference, Figures 8 to 10 are response waveforms, and Figure 11 is for other FIG. 12 is a configuration diagram showing an embodiment of the present invention, FIG. 12 is a configuration diagram of a conventional controller, and FIGS. 13 and 14 are diagrams explaining its operation. DESCRIPTION OF SYMBOLS 1... l'ID control calculation part, 2... Process, 11
... Auxiliary control section, 12... Selection section, 13... Tuner. Figure 1 111 Yutsuba Figure 3 Figure 4 (A) (B) Figure 5 (A) (B) Figure 6 Membership function Figure 9 Figure 10 (A) (B) Time - Time and Figure 12 Figure 13 Figure 1 ÷ Time Figure 14 (A) (B) (C) (D) -110 Hours Procedure Manual (Method) % Formula % Name of Invention Controller Name Yokogawa Electric Co., Ltd. agent 6, column 7 for a brief explanation of drawings in the specification subject to amendment, and characteristic curve diagrams enlarged with the contents of the amendment, Figures 5 to 7 are amended as follows.゛ 0 ”

Claims (6)

[Claims] (1) A control calculation section into which the target set value or the set value and the process amount are input and outputs the manipulated variable to the process by a predetermined calculation; and an auxiliary control section into which the target set value or the set value and the process amount are input. The auxiliary control section monitors the situation in which the process amount settles to the target set value or the set value, calculates the auxiliary set value, and outputs it to the control calculation section to prevent overshoot of the process amount. A controller characterized in that it prevents (2) A comparison value between the amount of change in the process amount per predetermined time and the target set value or the deviation of the process amount from the set value is used as an index for monitoring overshoot of the process amount. The controller according to claim 1. (3) The auxiliary setting value is moved closer to or away from the target setting value or the setting value by referring to an index for monitoring overshoot of the process amount. Controller. (4) When the target set value, set value, and auxiliary set value are mixed with different values, the set value or the auxiliary set value, whichever is farther from the target set value, is output to the control calculation section. The controller according to claim 1, characterized in that: (5) The controller according to claim 2, wherein an integral time value or a differential time value set in a control calculation section is used as the predetermined value. (6) The controller according to claim 2, wherein a value proportional to the equivalent dead time of the process is used as the predetermined value.