JPH0610765B2 - Process control equipment - Google Patents

Description

TECHNICAL FIELD OF THE INVENTION The present invention relates to an improvement of a process control device including a feedback control system and a feedback control system,
In particular, the present invention relates to a process control device that performs optimum control by varying a process transfer function according to actual process data.

[Technical background of the invention]

A conventional device of this type usually has a configuration in which a feedforward control system for the disturbance 1 is added to the feedback control system as an auxiliary as shown in FIG. 3, and the mainstream of the control is feedback control by the PID controller 2. It was due to. Specifically, the control value PV output from the process 3 having the transfer function G _P is fed back to the deviation calculation element 4, where the deviation between the control value PV and the target value SV is obtained, and this deviation is adjusted by PID. A feedback control system that controls the process 3 by performing a PID calculation by a total of 2 to obtain an operation output value MV that approaches zero, and the disturbance 1
As a compensation means in, the feedforward transfer function G _C for disturbance compensation is added to the output of the PID controller 2 and the process 3
And a feedforward control system in which a transfer function G _N for the disturbance 1 is added to the output side of.

[Problems of background technology]

By the way, in the above-described process control device, the feedback control is the mainstream, and the parameter that becomes the transfer function G _{P of the} process 3 used for the feedforward control is not a method that can be automatically identified during the control of the process 3. However, it is necessary for a person to separately calculate and measure actual process data and obtain and set each parameter, which makes the operation complicated and the control responsiveness deteriorates.

On the other hand, if the state equation of the plant is clear and all state variables can be easily measured, it is considered that a complete process controller can be realized, but in reality it represents a clear state equation of the plant. It is difficult to measure all state variables, and it is impossible to execute optimum control using the conventional device.

[Object of the Invention]

The present invention has been made paying attention to the above points. The transfer function of the process is sequentially changed to an optimum value according to the state of the process, the feedforward control is executed, and the fine adjustment is performed by the feedback control. It is an object of the present invention to provide a process control device that eliminates the influence of stability and delay time of process control.

[Outline of Invention]

The present invention measures the process gain in a stable state of the process, updates the process gain, obtains and corrects the process time constant from the change state at the time of transient response, and when the feedforward control system side is switched. It is a process control device that executes feedforward control while varying the process transfer function to an optimum value by measuring the time from when the control value changes to updating the dead time of the process.

Example of Invention

An embodiment of the present invention will be described below with reference to FIGS. 1 and 2. The transfer function G _{P of a} process is not as simple as a mathematical expression, but usually Can be approximated by By the way, in the process control device, the shortest rise time T _S with the minimum process dead time L shown in FIG.
When the control value PV is set to the target value SV in Fig. 1, Fig. 1 (A)
It is necessary to control the process by the operation output value MV as shown in. That is, as the operation output value MV, the process is controlled using the maximum operation output value MH during the rising time T _S = m · _Δt , and then, (However, if K is the process gain), the control value PV does not change like the dotted line a shown in FIG. 1 (B) to cause the hunting phenomenon, and the target value SV is accurately set. Can be reached. For this purpose, it is necessary to change necessary parameters T (process time constant), K (process gain) and L (process dead time) of the process transfer function G _P according to the state of the process during feedforward control. .

An embodiment of the process control device according to the present invention realized under the above demand will be described below with reference to FIG. In the figure, 11 is a deviation calculation means for calculating a deviation e between the control value PV obtained from the control process 12 and the target value SV, and the deviation e obtained here is used in addition to the PID adjusting section 13 as a feedback control system. A maximum operation output value output means 14 for outputting the maximum operation output value MH,
It is introduced into a deviation holding circuit 15 that always holds the previous deviation e ′, a control change value calculation means 16 that subtracts the previous deviation e ′ from the current deviation e to obtain a control change value h, a process time constant correction means 17, and the like. .

Reference numeral 18 denotes a gain updating means, which obtains the process gain K from the ratio between the operation output value MV and the control value PV when the control value SV and the target value SV are stable for a predetermined time or longer. The gain is updated, and the update process gain K is supplied to the predetermined operation output value output means 19.
The output means 19 receives the target value SV and outputs an operation output value SV / K shown in FIG. 1 (A). The output terminals of the maximum operation output value output means 14 and the predetermined operation output value output means 19 are selected by the first selection circuit 21 controlled by the maximum operation output value output time calculation means 20. The maximum operation output value output time calculation means 20 calculates the time T _s = m · for outputting the maximum operation output value MH by calculation using both outputs from the control change value calculation means 16 and the process time constant correction means 17. It has a function of obtaining _Δt , selecting the maximum operation output value output means 14 via the first selection circuit 21 during that time, and then selecting the predetermined operation output value output means 19. Reference numeral 22 denotes dead time updating means, which is the dead time L until the second selection circuit 23 switches from the feedback control system to the feedforward control system side, that is, the first selection circuit 21 side, and the control value PV changes. And has a function of closing the second selection circuit 23 on the feedforward control system side during L + Ts.

Next, the control value P shown in FIG.
The case of obtaining V will be described. PID controller 13 at all times
The operation of process 12 is controlled by
When the target value SV is set at the shortest rising time Ts as shown in (B), when the deviation e between the control value PV and the target value SV becomes a certain value or more, or the second selection circuit 23 is appropriately selected. And switch to the feedforward control system side. At this time, the deviation before switching the feedforward control system is already held in the deviation holding circuit 15 as the previous deviation e '.
In this state, if the deviation e after switching the feedforward control system becomes a predetermined value or more, the maximum operation output value output means 14
The maximum operation output value MH is output from the first selection circuit 2
1 to the process 12. On the other hand, the current deviation e is supplied to the control change value calculating means 16, where the control change value h = e−e ′, which is the difference between the current deviation e and the previous deviation e ′ of the deviation holding circuit 15, is obtained. This control change value signal h is sent to the maximum operation output value output time calculating means 20. The maximum operation output value output time calculation means 20 is supplied with the correction time constant T = T '+ Be for the current deviation e from the process time constant correction means 17. T'is a time constant of the previous deviation e ', and B is a constant determined by the accuracy of the process transfer function. When the control change value h = e-e 'and the correction time constant T = T' + Be are thus received, the maximum operation output value output time calculating means 20 causes the maximum operation output value output time shown in FIG. The time Ts is calculated by the following formula.

Then, during this time Ts, the first selection circuit 21 selects the maximum operation output value output means 14 side. The output time Ts can be derived based on the calculation formula described below. Without losing _{Ts = m · Δ T ...... (} 4) becomes equation is satisfied in a process control system. However,
K is a process gain, h is a control change value, and Δ _T is a predetermined control cycle.

Then, by transforming equation (2), MH ・ K−MH ・ Kd ^m = 1 MH ・ Kd ^m ＝ MH ・ K-1 And this is Can be expressed as Substituting equation (2) into equation (6), Becomes Furthermore, substituting equation (3) into this equation, Is established. Therefore, substituting this m into equation (4) gives (1)
The formula holds. Therefore, the optimum maximum operation output value output time Ts is determined based on the correction time constant T and the control change value h, and the maximum operation output value MH is given to the process 12 during this period.

Then, after the maximum operation output value output time Ts has elapsed, the first selection circuit 21 is connected to the predetermined operation output value output means 19 side. This means 19 obtains a predetermined operation output of SV / K, but this process gain K is a value updated by the PV / MV ratio when the process is stable when PVSV continues for a certain time or longer. It Therefore, the predetermined operation output value output means 19 gives the predetermined operation output value SV / K to the process 12.

On the other hand, the dead time updating means 22 measures the time from the switching to the feedforward control system until the control value of the process 12 changes, and considers the delay time Ts of the response to the second selection circuit 23 during that time. Selects and holds the feedforward control system side.

In this way, the process constants used in the feedforward control system are identified moment by moment, and the feedforward control is performed using the optimum parameters without hunting. It should be noted that when the feedback control is operated during the operation of the feedforward control, an overshoot occurs. Therefore, in order to prevent this, the operation by the feedback control is not performed during the operation of the feedforward control.

However, even when the feedforward control is performed, a control deviation occurs due to unclear plant characteristics. Therefore, the second selection circuit 23 switches to the feedback control system to perform the feedback control to perform the fine adjustment control of the process 12. Done.

Therefore, according to the above configuration, the transfer function of the process Among them, the process is approximated by the first-order lag system by the parameters of K, T, and L, and the feedforward control is performed. At the same time, the deviation by the process parameter is finely adjusted by the feedback control by PID. The control value can be set to the target value without causing a phenomenon. In particular, in a feedback control system when there is a process dead time L, it is necessary to wait for the control until a response appears in the control value, but this is not necessary in the case of feedforward control, and the dead time is expected in advance for the control. It can be applied, and the enactment time can be shortened accordingly.

In the above embodiment, the process transfer function is The feedback control system by PID control was used for the fine adjustment, and the sample PI, I-PD was used in place of this PID control.
It is also possible to control such as. Further, the process control device shown in FIG. 2 may include a control system in which feedforward control for disturbance is added. Also, the process transfer function T, K, and L may be functions of the control value, the operation output value, or the target value, respectively. In addition, the present invention can be modified in various ways without departing from the scope of the invention.

〔The invention's effect〕

As described in detail above, according to the present invention, the feedforward control is performed while the process transfer function is sequentially changed by measuring the process gain in the stable state and identifying the process from the deviation at the transient response. Since the control delay due to the dead time of the process can be eliminated and the hunting phenomenon can be eliminated, and the fine adjustment by the feedback control is added after the feedforward control, it is possible to provide the process control device capable of ensuring the stability of the process.

[Brief description of drawings]

FIGS. 1 and 2 are for explaining one embodiment of the process control device according to the present invention. FIG. 1 is a diagram showing a relationship between an operation output value and a process control value, and FIG. FIG. 3 is a schematic configuration diagram of the apparatus, and is a schematic view of a conventional process control apparatus including a feedforward control system expressed by using a transfer function. 11 ... Deviation calculation means, 12 ... Process, 13 ... PID adjustment part, 14 ... Maximum operation output value output means, 15 ... Deviation holding circuit, 16 ... Control change value calculation means, 17 ... Time constant correction means, 18 ... Gain update Means, 19 ... Predetermined operation output value output means, 20 ... Maximum operation output value output time calculation means, 22 ...
Dead time update means.

Claims (1)

[Claims] 1. A feedback control system of a process and a feedforward control system for preventing a hunting phenomenon of a control value are switchably provided, and the feed is provided when at least a deviation between the control value and a target value is a predetermined value or more. In a process control device for controlling the process by switching to a forward control system side, the feedforward control system obtains a correction process time constant T based on a deviation of this time obtained by a control value and a target value. And a deviation signal obtained by the time constant correction means and a difference signal between the present deviation after the switching and the previous deviation before the switching when the deviation after the switching to the feedforward control system side is a predetermined value or more. The maximum operation output value output time is calculated using the process time constant, and the maximum operation output value output time A first operation output value output means for applying a maximum operation output value to the process, the maximum after the operation output value output time, process stability during the new process gain K from the operation output value and the control value and the
And a second operation output value output means for applying a predetermined operation output value obtained by the process gain K and the target value to the process, and the control value of the process after switching to the feedforward control system side. And a dead time updating means for selectively holding the feed forward control system side during the dead time L of the feed forward control system and the delay time of the response of the process in consideration of the time until it changes. A process control apparatus, wherein feedforward control is executed while successively changing at least the parameters K, T and L of the process transfer function to optimum values during feedforward control.