JPH02264302A - Process control device - Google Patents

Abstract

PURPOSE:To eliminate the blank of feedback control, and to quickly settle control by making control arithmetic action function adequately according to the size of a dynamic characteristic compensation portion. CONSTITUTION:The process variable of controlled system 1 is detected by a detector 2, and difference from the output of a target value setting means 3 is supplied to a velocity type P-control means 22 and a velocity type PI-control means 23. The change of production quantity, etc., is detected by a disturbance quantity detecting means 11, and as for a static characteristic compensation portion, the difference from a preceding time is calculated by a position type/ velocity type signal converting means 13, and is supplied to an adding means 6. The dynamic characteristic compensation portion is incomplete-differentiated by a dynamic characteristic compensating means 14, and is supplied to the adding means 8, and simultaneously, a signal level discriminating means 15 switches a signal selecting means 23 according to its size. Namely, when an operation characteristic compensation portion is within a prescribed level, a velocity type PI-control means 21 is selected. When the operation characteristic compensation portion exceeds the prescribed level, the velocity type P-control means 22 is selected, and harmful I-operation is cut.

Description

[Detailed Description of the Invention] [Object of the Invention] (Industrial Application Field) The present invention relates to a process control device used in the instrumentation control system of various plants, and in particular, the present invention relates to a process control device used in the instrumentation control system of various plants. The present invention relates to a process control device that suppresses the influence of disturbances by operating PID wJ calculation in an appropriate state depending on the magnitude of dynamic characteristic compensation.

(Conventional technology) In recent years, production volume, load, etc. have fluctuated due to various demands.
One important theme is how to suppress and control the effects of these fluctuations at high speed, with high precision, and in a stable manner. In particular, from 1990 to the 21st century, users' demands for diversification, fluctuations in production volume, and higher quality have been increasing, and these demands will continue to increase. There is a strong desire to realize flexible automation.

Therefore, in conventional equipment, in order to satisfy the above requirements, a feedback (hereinafter abbreviated as FB) control system is used to control the process amount to be controlled so that it matches the target value, and a feedback (hereinafter abbreviated as FB) control system is used to satisfy the above requirements. This system is combined with a feedforward (hereinafter abbreviated as FF) control system that detects the amount of disturbance caused by one or more kinds of load changes and proactively provides a signal to cancel the influence of the disturbance. So-called F F/F B control is employed. That is, this FF/FB control employs a method of constantly adding and combining the output obtained by the FB control and the output obtained by the FF control to obtain a manipulated output to the controlled object.

By the way, the type of disturbance has the property of maintaining a constant amount for a certain period of time, increasing or decreasing at a certain point, and then maintaining the constant amount again, like the production amount (Figure 3 (a)) It can be divided into non-random disturbances such as , and random disturbances such as the amount of steam supplied to an unspecified load that changes completely randomly as shown in FIG.

In particular, this device targets non-random disturbances, such as the load (
Plate thickness.

These include changes in board width, line speed, etc.) and load on chemical heating furnace outlet temperature control (changes in production volume, etc.).

In such a process, if the output temperature is affected by changing the production volume, that is, the furnace load, the quality may directly deteriorate due to this effect, or the effect may propagate to downstream processes and affect the final product. Process disturbances can significantly reduce product quality, and when flexibly changing production volume in response to demand, or automatically changing plate thickness and width while maintaining the same line speed, process disturbances are Since the change occurs as shown in FIG. 3(a), a very important issue is how to suppress the excessive influence of this change.

Conventionally, in the case of non-random disturbances, when the dynamic characteristic compensation signal of FF control exceeds a predetermined level range,
B control is separated and only FF control is performed, and when the dynamic characteristic compensation signal is within a predetermined level range, rFF control + FB$l
This is handled by a combination of lJ.

FIG. 4 is a block diagram showing the configuration of such a conventional device. This control device detects the process amount from the controlled object 1 by the process amount detection means 2, and then introduces the detected process amount and the target value of the process amount set by the target value setting means 3 into the deviation calculation means 4. Then, the deviation calculation means 4 obtains a deviation signal e Q' -8Vn-PVn. This deviation signal el is the velocity type P1. It is guided to the i-node calculation means 5, where speed type PI adjustment calculation is performed based on the following formula, and ΔCn - kp 1 (en - en - 1) + (r/Tt
)・enl-(1) A speed type PI adjustment calculation output signal ΔCn is obtained.

In this formula, kp: proportional gain, Tl: integration time,
T: control period, en: current control deviation, en-1: previous control deviation.

On the other hand, in the FF control system, after the disturbance amount detection means 11 detects a disturbance m D n such as a load change, this disturbance H
The coefficient means 12 multiplies kDn by the feedforward gain k to obtain the feedforward Ha signal FFn.
After the feedforward control signal FFn is obtained, compensation is performed by dividing the feedforward control signal FFn into a static characteristic compensation system and a dynamic characteristic compensation system. This static characteristic compensation system uses the position type feedforward control signal F
Fn is converted by the position/velocity shadow signal conversion means 13 (F F n
−F F n−1) to calculate the speed shadow signal ΔF
After Fn conversion, the signal is input to the addition means 6, where the speed shadow signal and the speed shadow signal Δ from the speed shape P1 adjustment calculating means 5 are combined.
Cn to obtain 6M V n, which is further guided to the subsequent velocity/position signal converting means 7, and is converted into a position shadow signal M V n by performing the calculation MVn-MVn-1+ΔM V n have. - On the other hand, the dynamic characteristic compensation system transmits the position type feedforward control signal FFn to the dynamic characteristic compensation means 14.
The dynamic characteristic compensation signal Xn is obtained by incomplete differentiation at , and is introduced into the adding means 8 together with the output MVn of the velocity type/position type signal converting means 7.

Here, an example will be described in which the dynamic characteristic compensation signal Xn is obtained from the general formula of the FFFF control model p(s).

Now, assuming that the disturbance transfer function is G o (S) and the process transfer function is Gp (s), the FF control model G.

(s) is obtained by the formula G□(S) G p (S) = P(s) fine kII f(s) -k (1+ (f(s)-1)) It will be done. In equation (2), the underlined portion in the first half represents the static characteristic compensation, and the underlined portion in the second half represents the dynamic characteristic compensation. Therefore, the dynamic characteristic compensation signal Xn(s)
is, Xn(s) = 1f(s)-1) ・Dnx D
(S) ...(3) It can be determined by the following formula. The dynamic characteristic compensation signal Xn obtained in this manner is introduced into the adding means 8 as described above, and is added and synthesized with the position shadow signal M V n from the signal converting means 7 to obtain the operation signal Mvn'. is applied to the controlled object 1 to set the target value SVn and the process ff1Pv.
Control is performed so that n matches.

In addition, the dynamic characteristic compensation system is provided with a signal level determination means 15 on the output side of the dynamic characteristic compensation means 14, which compares the dynamic characteristic compensation signal Xn with a predetermined level signal δ, and determines whether δ<1Xnl When, the switch 9 is turned off, and when -δ≦Xn≦δ, the switch 9 is turned on.

In other words, when a non-random disturbance occurs, switch 9 is controlled off to use only FF control when dynamic characteristic compensation is acting strongly, and switch 9 is controlled on in a steady state where dynamic characteristic compensation is small (FF By combining the control and FB control, the influence of disturbances is suppressed.

(Problems to be Solved by the Invention) However, the following problems have been pointed out in the control device using the above-described disturbance compensation situation using the conventional FF control and the combination method of the FF control and the FB control.

(2) If a disturbance occurs due to a change in the load while a control deviation exists, and the dynamic characteristic compensation signal xn is larger than the predetermined level signal δ, the switch 9 is turned off and the FB control is disconnected. There is a problem with the control settling being delayed by that much.

■Also, if an undetectable disturbance or a change in the target value occurs while the FB control is disconnected due to the large dynamic characteristic compensation component, the control deviation will increase, and the deviation will be caused during the FF control. Even though we want to keep it at zero, that large deviation remains, and control settling is delayed in this respect as well.

In particular, in modern plants, the production volume is increasing and the production speed is increasing, and such excessive variations have a large impact on quality. Therefore, it is necessary to overcome such defects and create a device that can fully cope with the flexible automation era.

The present invention has been made in view of the above-mentioned circumstances, and by appropriately functioning the adjustment calculation operation according to the magnitude of the dynamic characteristic compensation, the blank of FB$11 is eliminated, control is quickly set, and controllability is improved. The objective is to provide a process control device that improves

[Structure of the Invention] (Means for Solving the Problems) In order to solve the above problems, the present invention provides a feedback control system with proportional/integral or proportional/integral/integral control based on the deviation.
a first adjustment calculation means that performs a differential adjustment calculation; a second adjustment calculation means that performs a proportional or proportional/differential adjustment operation excluding at least the integral based on the deviation; Selecting the output of the first adjustment calculation means upon receiving a signal in a steady state within a predetermined level range, and in a transient state in which the level of the dynamic characteristic compensation signal exceeds within the predetermined level range. In order to stop the integral/differential operation or the integral operation in response to the signal of
The configuration includes signal selection means for selecting the output of the adjustment calculation means.

Furthermore, the present invention has a configuration in which different proportional gains, which are optimum values, are set in the first adjustment calculation means and the second adjustment calculation means.

(Function) Therefore, by taking the above-mentioned measures, in the present invention, when a disturbance occurs and the dynamic characteristic compensation signal is in a transient state exceeding a predetermined level range, the signal selection means
Select the adjustment calculation means. Therefore, by performing at least the proportional or proportional/derivative adjustment operation excluding the integral operation, the second adjustment calculation means can reliably stop the integral operation that would be harmful in a transient state, and eliminate blanks in the FB control. Control settling can be accelerated.

In addition, by setting proportional gains suitable for the steady state and transient state in the first adjustment calculation means and the second adjustment calculation means, respectively, the deviation of the FB control does not become large during the transient state, and the controllability is improved. It can be improved.

(Example) Hereinafter, an example of the present invention will be described with reference to FIG. In this figure, the same parts as in FIG. 4 are given the same reference numerals, and a detailed explanation thereof will be omitted. Particularly different parts compared to the conventional apparatus will be explained below. That is,
In this device, in addition to the speed type PI or speed type PID (hereinafter collectively referred to as speed type PI) adjustment calculation means 21 on the output side of the deviation calculation means 4, the speed type P or speed type P
D (hereinafter collectively referred to as speed type P),
Both adjustment calculating means 2 are adjusted according to the magnitude of the dynamic characteristic compensation signal Xn obtained by the dynamic characteristic compensating means 14 when a disturbance occurs.
A signal selection means 23 is provided for selectively taking out the output of 1.22 and supplying it to the addition means 6.

That is, in this device, the signal selection means 23 selects the output of the velocity type PI:A@ calculation means 21 and supplies it to the addition means 6 in a steady state including the occurrence of disturbance, and also performs dynamic characteristic compensation against the occurrence of disturbance. The signal exceeded the predetermined level range. In a so-called transient state, at least the I operation or the ID operation, which has the effect of accumulating deviation, is stopped, and at the same time, the speed type P is set to take advantage of at least the P operation in order to eliminate the blank of FB$111.
The configuration is such that the output of the adjustment calculation means 22 is selected and supplied to the addition means 6.

Furthermore, this device has a speed type PlawB calculation means 21 and a speed type P adjustment calculation means 22 depending on the steady state and transient state.
In addition to being able to execute FBil;m1ll separately, the proportional gain of each adjustment calculation means 21, 22 is set to a different value, that is, an optimum value according to each state. That is, the optimal value of the proportional gain of FB, $-I1m in the transient state of disturbance compensation is different from the optimal value in the steady state, and the optimal value of the proportional gain will be described below.

Generally, the optimal value of the proportional gain is determined by the CHR method (K.

L., Chien, J.A., HronesS.J., B.

Re5w1ck's PID: J! After finding it using J [official] etc. and setting it as the initial value, the actual l1a is finally set.
The decision was made by conducting a small IRMI while looking at the IRMI characteristics.

Incidentally, the optimum proportional gain for disturbance suppression with no overshoot and minimum settling time according to the CHR method is as shown in the table below.

However, in the above table, T represents the process time constant, k represents the process gain, and L represents the dead time of the process. Above table
PID control or P! The optimum proportional gain of sub-control and P control is significantly different from 2 to 3 times.
Optimum control is achieved by individually setting proportional gains depending on the control.

Next, the operation of the apparatus configured as above will be explained.

The process Ik PV n from the controlled object 1 is detected by the process quantity detection means 2, and this process quantity PVn and the target value SVn of the process quantity output from the target value setting means 3 are guided to the deviation calculation means 4, where Deviation signal en m5Vn
- Obtain PVn.

This deviation signal en is supplied to the subsequent speed type PI adjustment calculation means 21.
and is sent to the speed type P adjustment calculation means 22. At this time, in a steady state, the signal selection means 23 is in the positional relationship as shown in the figure, so the speed type PI adjustment calculation means 21 is inserted in the FB control system. It will be. Therefore, this speed type PI adjustment calculation means 21 performs PI or PID adjustment calculation every control cycle to obtain the adjustment calculation output ΔCn, and then adds this adjustment calculation output ΔCn to the speed type/position type signal via the addition means 6. It is sent to the conversion means 7.

The signal conversion means 7 converts the velocity shadow signal into a position shadow signal, and then obtains a manipulation signal MV' via the addition means 8, and controls the controlled object 1 using this manipulation signal MV'.

By the way, when a disturbance occurs due to a change in production volume or a change in load, the disturbance amount detection means 11 detects the disturbance f.
After detecting fi D n, the coefficient means 12 multiplies this disturbance Q D n by a feedforward gain k to obtain a positional feedforward control signal FFn. After that,
This position-based feedforward control signal FFn is divided into a static characteristic compensation component and a dynamic characteristic compensation component.
is introduced into the position/velocity shadow signal conversion means 13 to obtain the difference between the previous value and the current value, that is, the velocity shadow signal ΔFFn, which is led to the addition means 6, where the velocity shadow signal ΔFFn and the velocity shape adjustment are obtained. Addition signal ΔM obtained by adding calculation output ΔCn
V n is supplied to velocity type/position type signal conversion means 7 .

On the other hand, in the latter operating characteristic compensation portion, the operating characteristic compensating means 1
4, the positional feedforward control signal FFn is incompletely differentiated, and the obtained incomplete differential value is.

That is, the dynamic characteristic compensation signal Xn is supplied to the signal level determining means 15 and the adding means 8. This signal level determination means 1
5, predetermined upper and lower limit level values ±δ are set in advance, and the dynamic characteristic compensation signal Xn and the predetermined upper and lower limit level values ±δ (δ>
0) and it is determined that the dynamic characteristic compensation signal selects the speed type P adjustment calculation means 22.

In other words, when -δ≦Xn≦δ, the signal selection means 23 selects (
(during a steady state) connect b and c; on the other hand, when δ x x n l (during a transient state) connect ac.

As a result, during the transient state, the summer operation that is harmful to the control of the transient state is stopped, and the speed type P adjustment calculation means 22
Or perform PD adjustment calculation to obtain adjustment calculation output △Cn,
This is sent to addition means 6. Moreover, since the optimum proportional gain is set according to the respective states of the speed type PI adjustment calculation means 21 which functions in a steady state and the speed type P adjustment calculation means 221 which functions in a transient state, there is no delay in control settling. Controllability can be greatly improved.

Therefore, according to the configuration of the embodiment described above, the compensation amount for the operating characteristics of the FF control is within a predetermined level range with respect to the occurrence of non-random disturbance. In the so-called steady state, there is a combination of rFF control plus FB control and PI or PID control, while the dynamic characteristic compensation bone is outside the predetermined level range.

In a so-called transient state, by controlling with the combination of "rFF@ control + FB control P or PD control", not only will there be no blank space in FB$ilI# during a transient state, but it will also be harmful to the control of a transient state! There is no movement, and control settling can be carried out quickly. Moreover, since the proportional gain of the FB control is set to different optimal values in the steady state and the transient state, controllability can be improved.

In the above embodiment, the responsiveness of the controlled object 1 is not particularly considered as a problem, but if the responsiveness is very slow, for example, a timer means 31 is installed on the output side of the signal level determining means 15 as shown in FIG. is provided, and when the dynamic characteristic compensation signal Xn enters from outside the predetermined level range to within the predetermined level range,
The timer means 31 determines the speed P in the transient state after a predetermined time has elapsed.
By controlling the signal selection means 23 so as to select the speed type PI adjustment calculation means 21 in the steady state from the adjustment operator stage 22, the FB control is utilized from the time when the FF control shows a complete response at the time of transition to the steady state. be able to.

In addition, the present invention can be implemented with various modifications without departing from the gist thereof.

C. Effects of the Invention] As explained above, according to the present invention, by appropriately functioning the adjustment calculation operation according to the magnitude of the dynamic characteristic compensation, control settling can be quickly performed while eliminating the blank of FB$1m. It is possible to provide a process $ control device that can perform the following steps and improve controllability.

4. rl! Brief explanation of J-plane FIG. 1 is a block diagram showing one embodiment of the process control device according to the present invention, FIG. 2 is a block diagram showing another embodiment of the present invention, and FIG. 3 is a block diagram showing the types of disturbances. The disturbance waveform diagram to be described, FIG. 4, is a configuration diagram of a conventional device.

1... Controlled object, 3... Target value setting means, 4...
Deviation calculation means, 6... Addition means, 7... Speed type/position type signal conversion means, 8... Addition means, 11... Disturbance amount detection means, 12... Coefficient means, 13...・Position type/
Speed shadow signal conversion means, 14...dynamic characteristic compensation means, 15
. . . Signal level discrimination means, 21 . . . Speed type PI adjustment calculation means, 22 . . . Speed type P, 31 clause calculation means, 23
. . . signal selection means, 31 . . . timer means.

Claims (2)

[Claims] (1) A feedback control system that extracts the deviation between the process quantity to be controlled and the target value using a deviation calculation means and performs adjustment calculations based on this deviation, and a feedforward control signal obtained by multiplying the disturbance amount by a feedforward coefficient. In the process control device, the feedback control system is configured to perform a proportional/integral or proportional/integral/derivative adjustment calculation based on the deviation. a second adjustment calculation means for performing a proportional or proportional/derivative adjustment operation excluding at least integral based on the deviation; Selecting the output of the first adjustment calculation means in response to a signal in a certain steady state, and performing at least integration in response to a signal in a transient state in which the level of the dynamic characteristic compensation signal exceeds the predetermined level range. A process control device comprising: signal selection means for selecting the output of the second adjustment calculation means in order to stop the operation. (2) The proportional gain of each of the first adjustment calculation means that performs adjustment calculations in a steady state and the second adjustment calculation means that performs adjustment calculations in an excessive state is set to a value that is optimal for each state. 1. The process control device according to 1.