阅读说明：本技术 一种面向无人值守小型压水堆的协调控制方法 (Coordination control method for unattended small pressurized water reactor ) 是由 成守宇 张博文 彭敏俊 于 2019-11-02 设计创作，主要内容包括：本发明公开了一种面向无人值守小型压水堆的协调控制方法,其特征在于,包括如下步骤：(1)建立协调控制器；(2)建立基于模糊逻辑的底层控制器；(3)建立基于神经网络PID的底层控制器；(4)协调控制方法实施,通过协调控制判断反应堆功率控制系统、稳压器压力控制系统和蒸汽发生器压力控制系统分别使用哪种底层控制器。本发明可以对小型压水堆进行不同于现有方式的协调控制,采用两种优势互补的底层控制器,提高了控制性能；能够更加准确地判断何种工况采用何种底层控制器,从而更大限度地发挥两种底层控制器的优势；能够通过协调控制器对底层控制器进行切换,来克服故障带来的影响。(The invention discloses a coordination control method for an unattended small pressurized water reactor, which is characterized by comprising the following steps: (1) establishing a coordination controller; (2) establishing a fuzzy logic-based bottom controller; (3) establishing a neural network PID-based bottom controller; (4) and (4) implementing a coordination control method, and judging which bottom layer controller is respectively used by the reactor power control system, the voltage stabilizer pressure control system and the steam generator pressure control system through coordination control. The invention can carry out coordination control different from the prior mode on the small pressurized water reactor, adopts two bottom controllers with complementary advantages and improves the control performance; the method can more accurately judge which working condition adopts which bottom layer controller, thereby exerting the advantages of the two bottom layer controllers to a greater extent; the impact caused by the fault can be overcome by switching the bottom controller through the coordination controller.)

1. A coordination control method for an unattended small pressurized water reactor is characterized by comprising the following steps:

(1) establishing a coordination controller, carrying out standardized processing on input parameters, fuzzifying input data, establishing a fuzzy rule base, carrying out fuzzy reasoning according to rules of the fuzzy rule base, and defuzzifying a reasoning result to obtain the output of the coordination controller;

(2) establishing a bottom controller based on fuzzy logic, simultaneously establishing a fuzzy rule table of the bottom controller, and giving an expression between input and output of the controller;

(3) establishing a neural network PID-based bottom layer controller, and giving an expression between controller input and controller output and a PID coefficient optimization method;

(4) and (4) implementing a coordination control method, and judging which bottom layer controller is respectively used by the reactor power control system, the voltage stabilizer pressure control system and the steam generator pressure control system through coordination control.

2. The coordinated control method for the unattended small pressurized water reactor according to claim 1, wherein the step (1) of establishing the coordinated controller specifically comprises the following steps:

a. input parameters are standardized: the input parameters comprise an actual reactor thermal power value and a set value, an actual coolant average temperature value and a set value, an actual pressure value and a set value of a pressure stabilizer, an actual feed water flow value and a set value, and an actual outlet steam pressure value and a set value of a direct-current steam generator;

b. fuzzifying input data: fuzzifying the standardized input data by adopting a Gaussian fuzzifier;

c. establishing a fuzzy rule base: respectively establishing rule bases for a reactor power control system, a steam pressure control system and a pressure stabilizer pressure control system;

d. fuzzy reasoning: in the fuzzy inference machine, an algebraic product operator is used as a T norm, and a maximum operator is used as an S norm;

e. and (3) deblurring: and adopting a central average defuzzifier, and according to the rule of a fuzzy rule base, when the bottom controller is selected to be the bottom controller based on fuzzy logic, the rule output is-1, and when the bottom controller is selected to be the bottom controller based on the neural network PID, the rule output is 1.

3. The coordinated control method for the unattended small pressurized water reactor according to claim 1, wherein the step (2) of establishing the fuzzy logic-based underlying controller specifically comprises the following steps:

a. establishing a fuzzy rule table of a bottom layer controller;

b. an expression of the controller output and input is given.

4. The coordinated control method for the unattended small pressurized water reactor according to claim 1, wherein the step (3) of establishing the neural network PID-based underlying controller specifically comprises:

a. giving an expression between controller input and output;

b. and (3) giving a weight coefficient optimization method: and optimizing the weight coefficient by adopting a gradient descent method.

5. The coordinated control method for the unattended small pressurized water reactor according to claim 1, wherein the coordinated control method in the step (4) is implemented and specifically comprises:

a. determining input and output of a coordination controller;

b. determining the input and output of the underlying controller based on fuzzy logic;

c. and determining the input and output of the underlying controller based on the neural network PID.

6. The coordinated control method for the unattended small pressurized water reactor according to claim 2, wherein the specific process of establishing the coordinated controller in the step (1) is as follows:

a. input parameters are standardized: the input parameters comprise an actual reactor thermal power value and a set value, an actual coolant average temperature value and a set value, an actual pressure value and a set value of a pressure stabilizer, an actual water supply flow value and a set value, and an actual outlet steam pressure value and a set value of a direct-current steam generator, and the standardization processing method is as follows:

in the formula, x_i-normalizing the pre-processing actual values;

x_i，min-a pre-normalization minimum;

x_i，max-a pre-normalization maximum value;

-normalized maximum value;

b. fuzzifying input data: adopting a Gaussian fuzzifier to fuzzify the standardized input data, wherein the Gaussian fuzzifier is a fuzzified mapping of the input data, and obtaining a membership function mu of a fuzzy set A' of input parameters after fuzzification_A′(x) As shown in the following formula

In the formula, a_i-describing the degree of dispersion of the data as a positive number;

x_i-the ith argument;

c. establishing a fuzzy rule base: establishing rule bases for a reactor power control system, a steam pressure control system and a pressure stabilizer pressure control system respectively to provide knowledge for fuzzy reasoning;

d. fuzzy reasoning: in the fuzzy inference machine, an algebraic product operator is used as a T norm, a maximum operator is used as an S norm, and a membership function of a jth output parameter on a corresponding fuzzy set is shown as the following formula

In the formula, σ_i-describing the degree of dispersion of the data as a positive number;

-fuzzy centroid of ith input parameter in l rule;

x_iP——

e. and (3) deblurring: using a central average deblurring device to obtain

in the formula (I), the compound is shown in the specification,the center of the l-th fuzzy set, i.e.

ω^lheight of the l-th fuzzy set, i.e. in the formula of step dThe coefficient of (a);

according to the fuzzy rule base rule, when the bottom layer controller selects FC,

7. The coordinated control method for the unattended small pressurized water reactor according to claim 3, wherein the specific process of establishing the fuzzy logic-based underlying controller in the step (2) is as follows:

a. establishing a fuzzy rule table of the bottom layer controller:

b. an expression is given for the controller output and input:

in the formula (I), the compound is shown in the specification,

8. The coordinated control method for the unattended small pressurized water reactor according to claim 4, wherein the specific process of establishing the neural network PID-based bottom controller in the step (3) is as follows:

a. given the expression between controller input and output: the neural network PID controller comprises a three-layer network, an input layer, a hidden layer and an output layer; the relationship between the input layer and the hidden layer input is shown as follows:

in the formula (I), the compound is shown in the specification,-a weight between the jth node of the input layer and the ith node of the hidden layer;

the relationship between the hidden layer input and the hidden layer output is shown as follows:

the relationship between the hidden layer output and the output layer is shown as follows:

in the formula (I), the compound is shown in the specification,

b. and (3) giving a weight coefficient optimization method: and (3) optimizing the weight coefficient by adopting a gradient descent method, wherein the weight coefficient is shown as the following formula:

in the formula (I), the compound is shown in the specification,

9. The coordinated control method for the unattended small pressurized water reactor according to claim 5, wherein the coordinated control method in the step (4) is implemented by the following specific processes:

a. determining the input and output of the coordination controller: the input parameters comprise an actual reactor thermal power value and a set value, an actual coolant average temperature value and a set value, an actual pressure value and a set value of a pressure stabilizer, an actual feed water flow value and a set value, and an actual outlet steam pressure value and a set value of a direct-current steam generator; when the output of the coordination controller is [ -1,0], selecting a bottom layer controller based on fuzzy logic, and when the output of the coordination controller is (0,1], selecting a bottom layer controller based on a neural network PID;

b. determining the input and output of the underlying controller based on fuzzy logic: for a reactor power control system, the steam flow deviation and the coolant average temperature deviation are input, and the control rod movement speed vector is output; for a steam pressure control system, the input is steam pressure deviation and steam pressure deviation change rate, and the output is water supply flow; for a pressure control system of the voltage stabilizer, the pressure deviation of the voltage stabilizer and the average temperature deviation of the coolant are input, and the electric heating power or the opening degree of a spray valve is output;

c. determining the input and the output of an underlying controller based on the neural network PID; for a reactor power control system, inputting the average temperature of the coolant and a set value thereof, and outputting a control rod movement speed vector; for a steam pressure control system, the input is steam pressure and a set value thereof, and the output is water supply flow; for the pressure control system of the voltage stabilizer, the input is the pressure of the voltage stabilizer and the set value thereof, and the output is the electric heating power or the opening degree of the spray valve.

Technical Field

The invention relates to a coordination control method, in particular to a coordination control method for an unattended small pressurized water reactor.

Background

The nuclear energy generated by the reactor core of the reactor is converted into heat energy under the slowing action of a coolant, the heat is transferred to a water working medium on the secondary side of the direct current steam generator through the coolant, the water is heated into steam, and the steam enters the steam turbine to do work to convert the heat energy into mechanical energy and electric energy. The nuclear safety is a life line of a nuclear power device, and after a target steady-state working condition is determined, an effective control method is required to ensure that the small pressurized water reactor is quickly, stably and safely transited to the target steady-state working condition from the current operating working condition, and particularly, the requirement on the control method is higher in a fault state.

The integrated bottom controller of the pressurized water reactor mainly comprises a reactor power controller, a pressure stabilizer pressure controller, a steam generator pressure controller and the like, and the bottom controllers are used for enabling controlled parameters to tend to set values; the coordination controller is used for coordinating the operation of each bottom layer controller, so that the parameter change in the transient process is more stable and safer.

At present, the research on a coordination control method of a small pressurized water reactor is less, and the method specifically comprises the following two steps:

(1) the method has the advantages that a coordination controller is not arranged, the matching of a primary loop and a secondary loop is ensured through a 'stack-following machine' operation mode, the set value of a bottom layer controller is generally kept constant, the same controller is adopted under any working condition, and the fault working condition is not considered;

(2) the coordinated controller is provided, the set value of the bottom layer controller is kept unchanged or linearly changed through a simple algorithm and rules, the same controller is adopted under any working condition, and the fault working condition is not considered.

In the method (1), because a coordinated controller is not provided, the operation mode of the 'stack-tracking machine' is difficult to adapt to the working condition of rapid variable load, and overshoot is easy to occur; in the method (2), although having a coordinated controller, the algorithm and rule are simple and are only used to change the set value of the controller, and optimization of the controller is not considered. The two methods described above have common problems: each bottom controller is internally provided with a control algorithm (such as a PID control algorithm), and the bottom controller cannot have a good control effect under all working conditions; the operation characteristics of the small pressurized water reactor can be changed under the fault working condition, and the original control method cannot be applied.

Disclosure of Invention

The invention aims to provide a coordination control method for an unattended small pressurized water reactor, and intelligent coordination control is realized.

In order to solve the technical problems, the invention adopts the following technical scheme:

a coordination control method for an unattended small pressurized water reactor comprises the following steps:

(1) establishing a coordination controller, carrying out standardized processing on input parameters, fuzzifying input data, establishing a fuzzy rule base, carrying out fuzzy reasoning according to rules of the fuzzy rule base, and defuzzifying a reasoning result to obtain the output of the coordination controller;

(2) establishing a bottom controller based on fuzzy logic, simultaneously establishing a fuzzy rule table of the bottom controller, and giving an expression between input and output of the controller;

(3) establishing a neural network PID-based bottom layer controller, and giving an expression between controller input and controller output and a PID coefficient optimization method;

(4) and (4) implementing a coordination control method, and judging which bottom layer controller is respectively used by the reactor power control system, the voltage stabilizer pressure control system and the steam generator pressure control system through coordination control.

Further, the establishing of the coordination controller in the step (1) specifically includes:

a. input parameters are standardized: the input parameters comprise an actual reactor thermal power value and a set value, an actual coolant average temperature value and a set value, an actual pressure value and a set value of a pressure stabilizer, an actual feed water flow value and a set value, and an actual outlet steam pressure value and a set value of a direct-current steam generator;

b. fuzzifying input data: fuzzifying the standardized input data by adopting a Gaussian fuzzifier;

c. establishing a fuzzy rule base: respectively establishing rule bases for a reactor power control system, a steam pressure control system and a pressure stabilizer pressure control system;

d. fuzzy reasoning: in the fuzzy inference machine, an algebraic product operator is used as a T norm, and a maximum operator is used as an S norm;

e. and (3) deblurring: and adopting a central average defuzzifier, and according to the rule of a fuzzy rule base, when the bottom controller is selected to be the bottom controller based on fuzzy logic, the rule output is-1, and when the bottom controller is selected to be the bottom controller based on the neural network PID, the rule output is 1.

Further, the establishing of the fuzzy logic-based bottom layer controller in the step (2) specifically includes:

a. establishing a fuzzy rule table of a bottom layer controller;

b. an expression of the controller output and input is given.

Further, the establishing of the neural network PID-based underlying controller in the step (3) specifically includes:

a. giving an expression between controller input and output;

b. and (3) giving a weight coefficient optimization method: and optimizing the weight coefficient by adopting a gradient descent method.

Further, the implementation of the coordination control method in step (4) specifically includes:

a. determining input and output of a coordination controller;

b. determining the input and output of the underlying controller based on fuzzy logic;

c. and determining the input and output of the underlying controller based on the neural network PID.

Further, the specific process of establishing the coordination controller in the step (1) is as follows:

a. input parameters are standardized: the input parameters comprise an actual reactor thermal power value and a set value, an actual coolant average temperature value and a set value, an actual pressure value and a set value of a pressure stabilizer, an actual water supply flow value and a set value, and an actual outlet steam pressure value and a set value of a direct-current steam generator, and the standardization processing method is as follows:

in the formula, x_i-normalizing the pre-processing actual values;

x_i，min-a pre-normalization minimum;

x_i，max-a pre-normalization maximum value;

-normalizing the processed actual value;

-normalized minimum value;

-normalized maximum value;

b. fuzzifying input data: adopting a Gaussian fuzzifier to fuzzify the standardized input data, wherein the Gaussian fuzzifier is a fuzzified mapping of the input data, and obtaining a membership function mu of a fuzzy set A' of input parameters after fuzzification_A′(x) As shown in the following formula

In the formula, a_i-describing the degree of dispersion of the data as a positive number;

x_i-the ith argument;

-the ith normalized input data;

c. establishing a fuzzy rule base: establishing rule bases for a reactor power control system, a steam pressure control system and a pressure stabilizer pressure control system respectively to provide knowledge for fuzzy reasoning;

d. fuzzy reasoning: in the fuzzy inference machine, an algebraic product operator is used as a T norm, a maximum operator is used as an S norm, and a membership function of a jth output parameter on a corresponding fuzzy set is shown as the following formula

In the formula, σ_i-describing the degree of dispersion of the data as a positive number;

-fuzzy centroid of ith input parameter in l rule;

x_ip——

-membership functions of the jth output parameter in the ith rule on the rule fuzzy set;

e. and (3) deblurring: using a central average deblurring device to obtain

The point at which the weighted center average is obtained is shown below:

in the formula (I), the compound is shown in the specification,

the center of the l-th fuzzy set, i.e.

Taking the maximum value of y_jA value;

ω^lheight of the l-th fuzzy set, i.e. in the formula of step d

The coefficient of (a);

according to the fuzzy rule base rule, when the bottom layer controller selects FC,at-1, when the underlying controller chooses BC,

is 1.

Further, the specific process of establishing the fuzzy logic-based underlying controller in the step (2) is as follows:

a. establishing a fuzzy rule table of the bottom layer controller: e.g. of the type₁And e₂Is an input parameter of the controller;

b. an expression is given for the controller output and input: e.g. of the type₁M language variable and e₂The nth linguistic variable corresponds to an output value of y_m，nThe controller output is shown as follows:

in the formula (I), the compound is shown in the specification,

——e₁the membership function corresponding to the mth linguistic variable;

——e₂the nth linguistic variable of (1) is associated with a membership function.

Further, the specific process of establishing the neural network PID-based underlying controller in the step (3) is as follows:

a. given the expression between controller input and output: the neural network PID controller comprises a three-layer network, an input layer, a hidden layer and an output layer; the relationship between the input layer and the hidden layer input is shown as follows:

in the formula (I), the compound is shown in the specification,¹ω_ij-a weight between the jth node of the input layer and the ith node of the hidden layer;

the relationship between the hidden layer input and the hidden layer output is shown as follows:

q₁(k)＝x₁(k)

q₂(k)＝q₂(k-1)+x₂(k)

q₃(k)＝x₃(k)-x₃(k-1)

the relationship between the hidden layer output and the output layer is shown as follows:

in the formula (I), the compound is shown in the specification,²ω_i-the weight between the ith node of the output layer and the hidden layer;

b. and (3) giving a weight coefficient optimization method: and (3) optimizing the weight coefficient by adopting a gradient descent method, wherein the weight coefficient is shown as the following formula:

in the formula, η 2 — the learning factor between the output layer and the hidden layer;

δ' (k) -the gradient between the output layer and the hidden layer;

η₁-a learning factor between the input layer and the hidden layer;

δ_i(k) between the input layer and the hidden layerOf the gradient of (c).

Further, the specific process implemented by the coordination control method in the step (4) is as follows:

a. determining the input and output of the coordination controller: the input parameters comprise an actual reactor thermal power value and a set value, an actual coolant average temperature value and a set value, an actual pressure value and a set value of a pressure stabilizer, an actual feed water flow value and a set value, and an actual outlet steam pressure value and a set value of a direct-current steam generator; when the output of the coordination controller is [ -1,0], selecting a bottom layer controller based on fuzzy logic, and when the output of the coordination controller is (0,1], selecting a bottom layer controller based on a neural network PID;

b. determining the input and output of the underlying controller based on fuzzy logic: for a reactor power control system, the steam flow deviation and the coolant average temperature deviation are input, and the control rod movement speed vector is output; for a steam pressure control system, the input is steam pressure deviation and steam pressure deviation change rate, and the output is water supply flow; for a pressure control system of the voltage stabilizer, the pressure deviation of the voltage stabilizer and the average temperature deviation of the coolant are input, and the electric heating power or the opening degree of a spray valve is output;

c. determining the input and the output of an underlying controller based on the neural network PID; for a reactor power control system, inputting the average temperature of the coolant and a set value thereof, and outputting a control rod movement speed vector; for a steam pressure control system, the input is steam pressure and a set value thereof, and the output is water supply flow; for the pressure control system of the voltage stabilizer, the input is the pressure of the voltage stabilizer and the set value thereof, and the output is the electric heating power or the opening degree of the spray valve.

Compared with the prior art, the invention has the beneficial technical effects that:

(1) the invention adopts two bottom controllers with complementary advantages, can adopt the bottom controller based on fuzzy logic when needing rapid regulation, and can adopt the bottom controller based on neural network PID when needing accurate regulation, compared with the bottom controller adopted under any working condition, the method improves the control performance.

(2) The coordination controller adopted by the invention is established based on the fuzzy logic theory, and can more accurately judge which working condition adopts which bottom layer controller, thereby exerting the advantages of the two bottom layer controllers to the maximum extent.

(3) Compared with the existing coordination control method which can not be suitable for the fault working condition, the coordination control method of the invention can obtain better performance under the fault working condition, and overcomes the influence caused by the fault by switching two bottom controllers through the coordination controller.

Drawings

The invention is further illustrated in the following description with reference to the drawings.

Fig. 1 is a flow chart of an intelligent coordination control method (an unattended small pressurized water reactor-oriented coordination control method) of the invention.

FIG. 2 is a schematic diagram of a rule base of a reactor power control system.

FIG. 3 is a schematic diagram of a steam pressure control system rule base.

FIG. 4 is a schematic diagram of a rule base of a pressure control system of a pressure regulator.

FIG. 5 is a schematic diagram of a fuzzy logic based rule base of an underlying controller.

FIG. 6 is a block diagram of a PID underlying controller flow based on a neural network.

FIG. 7 is a schematic diagram of a process implemented by the coordination controller under a specific operating condition.

Detailed Description

The invention is further described below with reference to the accompanying drawings.

With reference to fig. 1, the implementation steps of the present invention mainly include:

(1) establishing a coordination controller, which comprises the following specific processes:

a. input parameters are standardized: the input parameters comprise an actual reactor thermal power value and a set value, an actual coolant average temperature value and a set value, an actual pressure value and a set value of a pressure stabilizer, an actual water supply flow value and a set value, and an actual outlet steam pressure value and a set value of a direct-current steam generator, and the standardization processing method is as follows:

in the formula, x_i-normalizing the pre-processing actual values;

x_i，min-a pre-normalization minimum;

x_i，max-a pre-normalization maximum value;

-normalizing the processed actual value;

-normalized minimum value;

-normalized maximum value;

b. fuzzifying input data: adopting a Gaussian fuzzifier to fuzzify the standardized input data, wherein the Gaussian fuzzifier is a fuzzified mapping of the input data, and obtaining a membership function mu of a fuzzy set A' of input parameters after fuzzification_A' (x) is represented by the following formula

In the formula, a_i-describing the degree of dispersion of the data as a positive number;

x_i-the ith argument;

-the ith normalized input data;

c. establishing a fuzzy rule base: establishing rule bases for a reactor power control system, a steam pressure control system and a pressure stabilizer pressure control system respectively to provide knowledge for fuzzy reasoning;

d. fuzzy reasoning: in the fuzzy inference machine, an algebraic product operator is used as a T norm, a maximum operator is used as an S norm, and a membership function of a jth output parameter on a corresponding fuzzy set is shown as the following formula

In the formula, σ_i-describing the degree of dispersion of the data as a positive number;

-fuzzy centroid of ith input parameter in l rule;

x_ip——

-membership functions of the jth output parameter in the ith rule on the rule fuzzy set;

e. and (3) deblurring: using a central average deblurring device to obtain

The point at which the weighted center average is obtained is shown below:

in the formula (I), the compound is shown in the specification,

the center of the l-th fuzzy set, i.e.

Taking the maximum value of y_jA value;

ω^lheight of the l-th fuzzy set, i.e. in the formula of step d

The coefficient of (a);

according to the fuzzy rule base rule, when the bottom layer controller selects FC,

at-1, when the underlying controller chooses BC,

is 1.

(2) Establishing a bottom layer controller based on fuzzy logic, which comprises the following specific processes:

a. establishing a fuzzy rule table of the bottom layer controller: as shown in fig. 5: e.g. of the type₁And e₂Is an input parameter of the controller;

b. an expression is given for the controller output and input: see FIG. 5, e₁M language variable and e₂The nth linguistic variable corresponds to an output value of y_m，nThe controller output is shown as follows:

in the formula (I), the compound is shown in the specification,——e₁the membership function corresponding to the mth linguistic variable;

——e₂the nth linguistic variable of (1) is associated with a membership function.

(3) And establishing a neural network PID-based underlying controller. The specific process is as follows:

a. given the expression between controller input and output: as in fig. 6, the neural network PID controller includes a three-layer network, an input layer, a hidden layer and an output layer; the relationship between the input layer and the hidden layer input is shown as follows:

in the formula (I), the compound is shown in the specification,¹ω_ij-a weight between the jth node of the input layer and the ith node of the hidden layer;

the relationship between the hidden layer input and the hidden layer output is shown as follows:

q₁(k)＝x₁(k)

q₂(k)＝q₂(k-1)+x₂(k)

q₃(k)＝x₃(k)-x₃(k-1)

the relationship between the hidden layer output and the output layer is shown as follows:

in the formula (I), the compound is shown in the specification,²ω_i-the weight between the ith node of the output layer and the hidden layer;

b. and (3) giving a weight coefficient optimization method: and (3) optimizing the weight coefficient by adopting a gradient descent method, wherein the weight coefficient is shown as the following formula:

in the formula eta₂-a learning factor between the output layer and the hidden layer;

δ' (k) -the gradient between the output layer and the hidden layer;

η₁-a learning factor between the input layer and the hidden layer;

δ_i(k) -the gradient between the input layer and the hidden layer.

(4) The coordination control method is implemented by the following specific processes:

a. determining the input and output of the coordination controller: the input parameters comprise an actual reactor thermal power value and a set value, an actual coolant average temperature value and a set value, an actual pressure value and a set value of a pressure stabilizer, an actual feed water flow value and a set value, and an actual outlet steam pressure value and a set value of a direct-current steam generator; when the output of the coordination controller is [ -1,0], the bottom layer controller based on fuzzy logic is selected, and when the output of the coordination controller is (0,1], the bottom layer controller based on the neural network PID is selected.

b. Determining the input and output of the underlying controller based on fuzzy logic: for a reactor power control system, the steam flow deviation and the coolant average temperature deviation are input, and the control rod movement speed vector is output; for a steam pressure control system, the input is steam pressure deviation and steam pressure deviation change rate, and the output is water supply flow; for the pressure control system of the voltage stabilizer, the input is the pressure deviation of the voltage stabilizer and the average temperature deviation of the coolant, and the output is the electric heating power or the opening degree of the spray valve.

c. Determining the input and the output of an underlying controller based on the neural network PID; for a reactor power control system, inputting the average temperature of the coolant and a set value thereof, and outputting a control rod movement speed vector; for a steam pressure control system, the input is steam pressure and a set value thereof, and the output is water supply flow; for the pressure control system of the voltage stabilizer, the input is the pressure of the voltage stabilizer and the set value thereof, and the output is the electric heating power or the opening degree of the spray valve.

The actual coordinated control process of the small pressurized water reactor will be described below. The initial working condition is that the small pressurized water reactor stably runs under the power level of 60 percent FP, the average temperature of the coolant is 575.15K, the pressure of a pressure stabilizer is 15MPa, the water supply flow is 52.8kg/s, and the steam pressure is 3 MPa; the method comprises the following steps of (1) suddenly introducing all power failure faults of four main pumps, wherein the thermal power set value of a reactor is 20.5% FP, the average temperature set value of a coolant is 555.15K, the pressure of a voltage stabilizer is 15MPa, the water supply flow is 18.0kg/s, and the steam pressure is 3 MPa; in order to realize the stable transition of the process, the coordination control method provided by the invention comprises the following specific steps:

(1) and establishing a coordination controller. The input parameters were normalized as shown in the following table; in the process of fuzzification_iAnd σ_iIs 0.5; the rule base is shown in fig. 2, 3 and 4; calculating the fuzzy reasoning and defuzzification process according to a formula; fig. 7 illustrates the selection of the coordination controller given by the coordination controller for three control systems.

(2) And establishing a fuzzy logic-based bottom-layer controller. The rule base is shown in fig. 5, and the output of the underlying controller is calculated according to a formula by combining with the rule base.

(3) And establishing a neural network PID-based underlying controller. Weighting coefficient¹ω_ijAnd²ω_iis set to 1, the output of the controller is calculated according to a formula, and the weight coefficient is optimized on line.

(4) And implementing a coordination control method. Determining an input/output interface of a coordination controller; and determining input and output interfaces of the fuzzy logic-based bottom-layer controller and the neural network PID-based bottom-layer controller and three control systems.

The above-described embodiments are merely illustrative of the preferred embodiments of the present invention, and do not limit the scope of the present invention, and various modifications and improvements of the technical solutions of the present invention can be made by those skilled in the art without departing from the spirit of the present invention, and the technical solutions of the present invention are within the scope of the present invention defined by the claims.