JPH0355601A - Automatic adjustment method for control parameter - Google Patents

Abstract

PURPOSE:To attain the optimum adjustment of a control parameter so as to satisfy an index to plural process variables by obtaining the optimum control parameter of a controller after optimizing an evaluation function set previously. CONSTITUTION:An MG set speed controller 32 performs an arithmetic operation based on the deviation between the speed signal of an MG set generator 31 and the target value r(t) and outputs a speed control signal u(t) of the generator 31. At the same time, a process identifying part 34 inputs the output u(t), the MG set generator speed signal, a neutron flux serving as the adjustment specifications, and the main steam quantity signal and identifies the generator 31 as well as a reactor and a turbine 33. The process model identified by the part 34 is inputted to an optimum control parameter deciding part 35. The part 35 decides an optimum control parameter to set this parameter directly to the controller 32 or after confirmation of an operator.

Description

[Detailed Description of the Invention] [Object of the Invention] (Industrial Application Field) The present invention relates to a method for optimally automatically adjusting control parameters of a control device, and particularly relates to a method for automatically adjusting control parameters of a control device, and in particular, the present invention relates to a method for automatically adjusting control parameters of a control device, and in particular, the present invention relates to a method for automatically adjusting control parameters of a control device. The present invention relates to an automatic control parameter adjustment method suitable for optimizing control performance under constraint conditions such as control.

(Prior Art) Generally, in a control device that executes process control, it is necessary to optimally adjust the control parameters (for example, proportional gain, integral gain, etc.) of the control device. A method as shown in FIG. 5 is conventionally known as a method for adjusting.

That is, in FIG. 5, 1 is the controlled object (process), and 2 is the output y (t) of the controlled object 1 and the target value '(t)
This figure shows a control device that performs predetermined calculations (such as PID calculation) based on the deviation of .

Then, the process identification unit 3 inputs the input u (
t) and the output y (t), the transfer function model G p (s) of the controlled object 1 is identified, and the optimal control parameter determination unit 4 identifies the transfer function model G p (s) identified by the process identification unit 3. ) and a transfer function Gm(S) (reference model) indicating a desirable response given as a specification by a control system coordinator, etc., to derive optimal control parameters for the control device 2.

The optimal control parameters are derived by the optimal control parameter determining unit 4, for example, as described in "Control system design method based on partial knowledge of the controlled object" (Proceedings of the Society of Instrument and Control Engineers, Vol. 1).
5, No. 4, 1979), etc., as follows.

In other words, the target value r consisting of the controlled object 1 and the control device 2
The closed loop transfer function from (t) to the output y (t) is defined as W
(s), W(s) can be written as follows.

W(s) = (G c (s) G p (s)
1/ (1+(;c(s)Gp(s))...
-(1) On the other hand, the transfer function G m (s) of the reference model is given by the following equation.

Gm(s) = 1/(1+cy s +a2
((F S) ”+α3 (σs)゛1+...)
(2) Here, α2, α3, . . . are constants that govern the response shape, and are set in advance by the adjuster as control performance specifications of the control system. For example, in order to make the output y (t) stabilize without overshoot in response to a step change in the target value r0), α2
-0.375, α3-0.0625, .

By setting the closed-loop transfer function W(s) in equation (1) and the transfer function G m (s) of the reference model in equation (2) to be equal, the closed-loop control system can have a desired response. This is shown in the following equation.

W(s)-Gm(s) ・=- (3) G p
Since the coefficient of S (s) has been identified by the process identification unit 3, the unknown variables among the S coefficients in equation (3) are the control parameter of the controlled object G c (s) and σ of the reference model. As can be seen from equation (2), σ of the reference model is a parameter indicating the quick response. (3)
The identity for S in the equation is solved to obtain the optimum value of the control parameter that provides a desired response shape (for example, does not exhibit overshoot). However, by solving to minimize σ, it is possible to show the fastest possible response.

By the way, Figure 6 shows a boiling water nuclear power plant (BW).
In the BWR plant, steam generated in the reactor 11 passes through the main steam pipe 12 to the turbine 13, rotates the generator 14 via the turbine 13, and obtains electrical output. It is configured like this.

When heat generation in the reactor 11 is controlled by the recirculation flow system, a command signal 17 is sent from the MG set speed controller 15 to the MG set generator 16 to change the rotation speed of the MG set generator 16. .

Then, this change in the rotation speed changes the rotation speed of the motor 18 connected to the MG set generator 16, the rotation speed of the recirculation pump 19 driven by the motor 18, and the external recirculation flow rate 2.
0 and the core flow rate 21, and further, the change in the core flow rate 21 is transmitted to the void ratio in the core and the nuclear fission reaction, and the thermal output changes.

That is, in order to increase the output in a BWR plant, the neutron flux necessarily fluctuates, and the amount of this fluctuation is an important index when adjusting control parameters.

Therefore, when determining the optimal control parameters for the recirculation flow control system of a BWR plant, the optimal control parameters are determined so as to keep the fluctuations in neutron flux within the standard and to make the response of the plant output as fast as possible. However, the conventional methods described above cannot be used for the recirculation flow rate control system of such BWR plants because it is impossible to optimally adjust control parameters that satisfy indicators for multiple process variables. Ta.

(Problems to be Solved by the Invention) As described above, with the conventional control parameter automatic adjustment method, it is not possible to optimally adjust the control parameters so as to satisfy the indicators for multiple process variables. There was a problem that it could not be applied to a recirculation flow rate control system, etc.

The present invention has been made in response to such conventional circumstances, and is an automatic control parameter that can optimally adjust control parameters to satisfy indicators for multiple process variables: J! It is intended to provide a J4i method.

[Structure of the Invention] (Means for Solving the Problems) That is, the present invention identifies a process controlled by a control device using a mathematical model, and derives optimal control parameters for the control device based on this mathematical model. The control parameter automatic adjustment method is characterized in that the optimal control parameters are derived by optimizing a preset evaluation function.

(Function) In the control parameter automatic adjustment method of the present invention having the above configuration,
For example, in the recirculation flow control system of a BWR plant,
It is possible to optimally adjust control parameters to satisfy indicators for multiple process variables, such as determining optimal control parameters to keep neutron east fluctuations within standards and to make plant output response as fast as possible. can.

(Embodiment) Hereinafter, an embodiment of the control parameter automatic adjustment method of the present invention will be described with reference to the drawings.

FIG. 1 shows the configuration of an embodiment in which the present invention is applied to a recirculation flow rate control system of a BWR plant. In the figure, reference numeral 31 indicates the output u (t
) shows the MG set generator which is the controlled process, and the reference numeral 33 shows the nuclear reactor + Yuichi bin which is the controlled process.

The MG set speed controller 32 performs a predetermined calculation based on the deviation between the MG set generator speed signal and the target value "(t),"
Signal u for controlling the speed of the MG set generator 31
(t) is output.

Further, in the figure, reference numerals 34 and 35 indicate a process identification section and an optimal control parameter determination section, respectively.

This process identification unit 34 is connected to the MG set speed controller 32
output u (t) and the feedback signal MG
Input not only the set generator speed signal, but also the neutron flux and main steam flow rate signals, which are the adjustment specifications, and the process (M
G-set generator 31 and reactor + turbine 33) are identified.

The process model identified by the process identification unit 34 is input to the optimal control parameter determination unit 35, which determines the optimal control parameter by a method described later, and sets it directly to the MG set speed controller 32. Alternatively, after displaying it to the adjuster on a CRT or the like and obtaining confirmation from the adjuster, it is set in the MG set speed controller 32.

In the conventional method, the process model of the process identification unit 34 is limited to the transfer function in the S domain (which can be expressed by a rational expression of the Laplace operator S such as G p (s)), but in the present invention, the process model in the process identification unit 34 is Not only functions but also AR models or nonlinear models can be used as long as they are mathematical models that can be simulated in the time domain.

The optimal control parameter determination unit 35 derives optimal control parameters by optimizing the evaluation function. In this way, the calculation algorithm for optimizing the evaluation function is the simplex method, which is one of the nonlinear programming methods (for example, "Literature: Konno ◆ Yamashita, Nonlinear Programming,"
978J etc. ) can be used. Hereinafter, a method of automatically adjusting two control parameters, the proportional gain and the integral gain of the MG set generator speed controller 32, using the simplex method will be described.

FIG. 2 shows an overview of the calculation algorithm for determining the optimal control parameters based on the simplex method. below,
Steps 1 to 6 in FIG. 2 will be explained step by step.

Step 1: Initial Simplex Setting In the simplex method, one parameter set is prepared that is one more than the number of control parameters to be adjusted. In this example, there are two control parameters to be adjusted, a proportional gain and an integral gain (n-2), so three combinations x1 of (proportional gain, integral gain) are prepared (1-1.2,... .n
+1). This parameter set Xi (1-1.2.
・・・． n+1) is called a simplex. These three groups are [
Proportional gain - Integral gain] corresponds to each vertex of a triangle on the plane.

Step 2: A closed-loop system simulation is performed at each point of the three sets (triangle vertices) xi (1=1.2...n+1) set in step 1.

This simulation is performed using the process model identified by the process identification unit 34 and the model of the MG set speed controller 32 (a nonlinear model is also possible). Figure 3 shows an example of the response obtained as a result of the simulation.
Three sets of such simulation results are obtained. Note that the symbols a1b and c in the figure indicate the responses of the MG set generator speed, neutron flux, and main steam flow rate, respectively, and the overshoot amount (OVR), response time (Tr), etc. are characteristic values that serve as adjustment standards. It is.

Step 3: From the response obtained in the simulation in Step 2, characteristic values that serve as adjustment standards for the overshoot amount of the neutron flux, the response time of the main steam flow rate, etc. are calculated.

Step 4: Calculate the evaluation function ψ(Xi) from the response characteristic values obtained in Step 3.

2 ψ(xi) − ``(xi) +Σ ψ (gj(x
i). gmaxj)j-1 (j=1.2.n+1)... (4) Here, 1'(X); Main steam flow rate response time T ``ψ(g.g*ax); O, g<
gmax; v*(g-gmax)2+ g > go
+ax... (5) g+ (x); Neutron east overshoot OVRg
z (x); The neutron east attenuation ratio ψ (g.gmax) provides an index that becomes a constraint during parameter adjustment as a penalty for the objective function r(x).

In equation (5), no penalty is applied when the limit value glaX is not exceeded, but a penalty is applied when the limit value gsax is exceeded. Furthermore, the penalty weighting coefficient V is set to a sufficiently large value. Neutron overshoot and neutron attenuation ratio were set as constraint conditions.

The vertex xhSxsSxl is defined as follows depending on the size of the evaluation function.

xh; vertex xs that gives the largest ψ(xi) among the three vertices
;The vertex that gives the second largest ψ(xl) among the three vertices! I: The vertex that gives the smallest ψ(x1) among the three vertices.Furthermore, define the centroid (X) generated from vertices other than xh. In the case of rr2, it becomes the midpoint between (X> and xl.

<X) − (1/n) Σx1 −
...- (6)l≠h Step 5; Evaluation function ψ(xi
) to determine convergence. For example, the convergence determination is performed using the following equation.

ni1 (1/(n+1)Σ (ψ−ψ(xl))”l” <
e1-1 ... (7) Here, ni1 φ set (1/n+1) Σ ψ (xi) ...
(8) 1-1 When convergence determination formula (7) is not satisfied, step 6 is performed. Steps 2 to 6 are repeated until this determination formula is satisfied. When the determination formula is satisfied, the calculation ends, and xi, which takes the minimum ψ(Xi) at that time, becomes the optimal parameter combination.

Step 6; Modify simplex xi. The basic pattern of change is as follows.

Reflection: Create a reflection point X' from <x> and xh.

xr- (1+a)<x>-a xh
a > 0 expansion; Extend in the xr direction.

xe= γxr+(1-7)(X:) 7
>1 contraction; xh is contracted in the <1B> direction.

Xe−βxh+ (1−β)<X>0<β×l reduction: All vertices are reduced in the direction of x1.

Xi- (1/2) (xi+ xl) i-1.
・．． n+1 An example of a simbrex change algorithm using the above basic change pattern is shown in FIG.

As described above, according to this embodiment, the optimal control parameters can be easily determined even in the BWR recirculation flow rate control system where the adjustment criteria are complex and highly nonlinear, and thereby the control parameters can be automatically adjusted. This makes it possible to shorten control parameter adjustment time and improve control performance of systems that are difficult to adjust.

[Effects of the Invention] As explained above, according to the present invention, for example, BWR
In the recirculation flow control system of a plant, we aim to satisfy indicators for multiple process variables, such as determining optimal control parameters to keep neutron flux fluctuations within standards and to make plant output response as fast as possible. Optimal adjustment of control parameters can be made.

[Brief explanation of drawings]

FIG. 1 is a diagram for explaining the configuration of an embodiment of the automatic control parameter adjustment method according to the present invention, FIG. 2 is a flowchart for explaining an overview of the method for determining optimal control parameters according to the present invention, and FIG. Figure 4 is a diagram to explain the adjustment evaluation index in the step response of the recirculation flow rate #J system, Figure 4 is a diagram to explain the simplex change algorithm of the simplex method used in the embodiment of the present invention, Figure 5 6 is a diagram for explaining a conventional control parameter automatic adjustment method, and FIG. 6 is a diagram for explaining a BWR recirculation flow rate control system. 31...MG set generator 32...MG set speed controller 33...
・Nuclear reactor + turbine 34...Process identification section

Claims (2)

[Claims] (1) In a control parameter automatic adjustment method in which a process to be controlled by a control device is identified using a mathematical model, and optimal control parameters of the control device are derived based on the mathematical model, the optimal control parameters are set in advance. A control parameter automatic adjustment method characterized by optimizing and deriving an evaluation function. (2) The control device is a recirculation flow rate control device for a boiling water nuclear power plant, and the evaluation function has a plant output response time as an objective function, and a neutron overshoot amount and a neutron flux attenuation ratio as constraints. 2. The control parameter automatic adjustment method according to claim 1, wherein: