A design method for optimal PID parameters of closed-loop control systems in nuclear power plants
By using genetic algorithms to identify the controlled object model and optimize PID parameters on the nuclear power plant design verification platform, the problem of relying on empirical data on the parameter setting of the nuclear power plant closed-loop control system is solved, and the stable operation of the system and parameter availability are improved.
Patent Information
- Application Number
- CN202011341564.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-11-25
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2040-11-25
AI Technical Summary
In nuclear power plants, the parameter setting of the closed-loop PID control system mainly relies on empirical data and lacks theoretical basis, which increases the workload of on-site personnel and the risk of system operation instability during debugging.
Using a genetic algorithm-based method, a nuclear power plant design verification platform is used for simulation. By identifying the controlled object model and optimizing PID parameters, the stable operation of the closed-loop control system is ensured and the engineering index requirements are met.
It provides effective theoretical calculation basis, improves the availability of PID parameters, reduces the on-site debugging workload and system instability risks, and is suitable for the optimization design of closed-loop control systems in nuclear power plants.
Smart Images

Figure CN112666821B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of application of intelligent algorithms in nuclear industry, and in particular relates to a design method for optimal PID parameters of a closed-loop control system of a nuclear power plant. Background Art
[0002] With the development of China's nuclear power technology, the requirements for the safety and reliability of nuclear power operation have gradually increased, and higher requirements have been put forward for the degree of automation of process control in nuclear power plants, which include the safety, reliability and stability of control, as well as the accuracy of control results. There are many closed-loop control systems in nuclear power plant systems, and the closed-loop PID control method is used to ensure the automatic, safe and stable operation of relevant systems in nuclear power plants.
[0003] In previous project designs, the PID parameters (proportional P, integral I, differential D) of the PID controller of the nuclear power plant closed-loop control system were basically set based on empirical data (PID controllers achieve control functions by setting PID proportional, integral, and differential parameters), and the empirical data mainly came from the French reference power plant. For our independently designed Hualong One reactor-type nuclear power plant, the PID parameter setting of the closed-loop control system is a very important part of the design process. If there is no certain theoretical basis and calculation basis, the PID parameters will be debugged and adjusted during the commissioning and operation of the nuclear power plant, which will increase the workload of on-site personnel and the risk of system operation instability.
[0004] With the development of nuclear power technology, each nuclear power project has built a nuclear power plant design verification platform, which uses the mechanism modeling method to build a full-range engineering simulator for nuclear power plants. Through the real logic of the nuclear power plant instrumentation and control system configured in the workstation and server, the commands issued by the design verification can be automatically processed in the server, issue instructions to the equipment and give feedback, just like the automation system of a real nuclear power plant, to verify the function of the instrumentation and control system. Summary of the invention
[0005] The purpose of this invention is to propose a design method for the optimal PID parameters of a nuclear power plant closed-loop control system based on a genetic algorithm based on the current status of nuclear power technology and the use of nuclear power plant design verification platform simulation data, so as to correctly obtain the optimal PID control parameters suitable for nuclear power plants and ensure the optimal control of nuclear power closed-loop control systems.
[0006] To achieve the above purpose, the technical solution of the present invention is to provide a design method for optimal PID parameters of a closed-loop control system of a nuclear power plant, which is used to obtain the optimal PID parameters of a closed-loop control system that meets engineering index requirements, comprising the following steps:
[0007] Step S1, determining a closed-loop control system whose PID parameters need to be optimized, drawing a closed-loop control system block diagram of the closed-loop control system, and determining a controlled object;
[0008] Step S2, performing a nuclear power plant simulation on a nuclear power plant design verification platform and running the closed-loop control system;
[0009] Step S3, determining the model structure and model parameters of the controlled object, identifying the controlled object using a genetic algorithm, obtaining the optimally identified model parameters, and acquiring the final controlled object model;
[0010] Step S4, using the genetic algorithm to optimize PID parameters and obtain the optimal PID parameters;
[0011] Step S5, using mathematical software to substitute the controlled object model and the optimized PID parameters into the closed-loop control system block diagram for simulation until the optimized PID parameters meet the engineering index requirements for stable operation of the closed-loop control system;
[0012] Step S6, writing the optimized PID parameters that meet the engineering index requirements for stable operation of the closed-loop control system into an engineering design document.
[0013] Further, in the step S1, a closed-loop control system whose PID parameters need to be optimized is determined, and a block diagram of the closed-loop control system is drawn according to the process operating conditions, the controlled object is determined, and the controlled object control input u and the system response output y are clarified. The closed-loop control system block diagram includes a PID controller, an actuator and a first controlled object, and the PID controller, the actuator and the first controlled object constitute a closed-loop structure model.
[0014] Further, in the step S2, a nuclear power plant simulation is performed on a nuclear power plant design verification platform to operate the closed-loop control system. When the closed-loop control system is in a steady-state operation stage, an excitation signal is stimulated. The excitation signal can be stimulated at the control input end of the controlled object, can be stimulated by changing the fixed value mode of the closed-loop control system, and can also be stimulated by external disturbances. After the excitation signal is triggered, the closed-loop control system gradually transitions from an initial steady state to a new steady state, and records the process change values of the controlled object control input u and the system response output y.
[0015] Further, in the step S3, the model structure of the controlled object is determined according to the mechanism characteristics of the controlled object and the form of the system response output y, the model parameters that need to be identified are determined according to the model structure of the controlled object, and the controlled object is identified using a genetic algorithm; when the identification result meets the first evaluation index, the optimally identified model parameters are obtained; the first evaluation index refers to the minimum deviation between the system response output of the controlled object and the response output of the identification system, the output of the system response output of the controlled object and the response output of the identification system refers to the output signal generated when the input signal excites the controlled object system or the identification system, and the output signal changes over time as an output curve.
[0016] further,
[0017] In step S3, the controlled object model structure includes the following:
[0018] Expanded form
[0019] Zero-pole form
[0020] Multi-capacity inertia link
[0021] First-order system plus pure delay
[0022] Other forms
[0023] G(s) is the transfer function of the controlled object model
[0024] K is the proportionality factor
[0025] T 1…n is the time coefficient
[0026] e -τs It is a delay phase
[0027] n is the transfer function order
[0028] Determine the model parameters (KT) that need to be identified according to the model structure of the controlled object 1 T 2 …T n t 0 t 1 …t n-1 τ), using the genetic algorithm to identify the controlled object; determining the first evaluation index of the genetic algorithm by formula (6): comparing the controlled object response output sequence y(k) with the identification model response output sequence y *The square sum of the errors of (k) is used as the objective function Q, so that the objective function Q reaches the global minimum; the objective function is selected to determine the fitness function as Q min ;
[0029]
[0030] y(k) is the output sequence of the controlled object obtained by the nuclear power plant design verification platform;
[0031] y * (k) is the output sequence of the controlled object identified by the genetic algorithm;
[0032] C 1 is a constant coefficient to ensure that Q min non-negative;
[0033] is the objective function, which ensures the minimum error of the identification result;
[0034] cov(abs(y(k)-y * (k)) is to suppress noise and ensure that the noise variance is minimized.
[0035] a 1 , b 1 is the error magnification factor to improve the recognition accuracy. 1 , b 1 , so that the identification result quickly reaches the global minimum;
[0036] Set the crossover factor between 0.4-0.9, the mutation factor between 0.01-0.1, and the population size between 30-60. min When it is the smallest, it indicates that the model parameter (KT 1 T 2 …T n t 0 t 1 …t n-1 τ) is optimal; the identified optimal model parameter (KT 1 T 2 …T n t 0 t 1 …t n-1 τ) is substituted into the corresponding controlled object model structure, that is, the final controlled object model is obtained, and the controlled object model includes the controlled object model structure.
[0037] Further, in step S4, the PID parameters are optimized using the genetic algorithm through a PID optimization program, the PID parameters found by the genetic algorithm are combined with the controlled object model structure identified by the controlled object, and a closed-loop control system model is calculated; the PID optimization program automatically stimulates a step response for the closed-loop control system; when the output curve of the closed-loop control system meets a second evaluation index, the optimized PID parameters are obtained; the second evaluation index refers to a control index of the quality of the response output curve of the closed-loop control system.
[0038] further,
[0039] In step S4, when the genetic algorithm is used to optimize the PID parameters, the closed-loop control system is first simplified into an open-loop system with an input of r and an output of y by using the basic simplification rule of the signal flow, wherein r is the closed-loop fixed value input quantity and y is the system response output y; a unit step is performed on r, and the step response output curve y of the closed-loop control system is subjected to the second evaluation index: the closed-loop fixed value step input sequence u(k) is compared with the closed-loop calculation model step response output sequence y * The square sum of the errors of (k) is used as the objective function J. When the step response output curve y does not meet the quality requirements, the objective function J is penalized by applying a penalty factor J', and finally the objective function J is made to reach the global minimum. The objective function is selected to determine the fitness function J min ; The step response output curve y is obtained by the change of the system response output y over time;
[0040]
[0041] u(k) is the closed-loop constant-value step input sequence;
[0042] y * (k) is the step response output sequence of the closed-loop calculation model;
[0043] C 2 is a constant coefficient to ensure that J min non-negative;
[0044] is the objective function, which ensures the minimum error of the identification result;
[0045] cov(abs(u(k)-y * (k)) is to suppress noise and ensure the minimum noise variance;
[0046] a 2 、b 2 is the error magnification factor to improve the recognition accuracy. 2 、b 2, so that the optimization result can quickly reach the global minimum;
[0047] J' is a penalty factor, which penalizes the step response output curve y that does not meet the engineering quality index;
[0048] The step response output curve y should meet the following quality index requirements as shown in formula (8):
[0049]
[0050] M p is the overshoot of the step response output curve y,
[0051] φ is the attenuation rate of the step response output curve y,
[0052] u pid is the output capability of the PID controller,
[0053] t s-pid It is the average time for the PID controller to change 100% opening;
[0054] The penalty factor J' can be adjusted and changed according to engineering requirements;
[0055] When the fitness function J min When it is the smallest, it indicates that the PID parameters obtained by the genetic algorithm are the optimal PID parameters. The optimal PID parameters are substituted into the PID controller in the closed-loop control system block diagram to obtain the final closed-loop structure model.
[0056] Further, in step S5, simulation verification is performed using mathematical software, the closed-loop control system block diagram is built, the controlled object model and the optimized PID parameters are substituted into the closed-loop control system block diagram for simulation; if the step response output curve characteristics of the closed-loop control system do not meet the engineering index requirements, the second evaluation index can be adjusted according to engineering requirements, and after the adjustment, the genetic algorithm is used again to optimize the PID parameters until the optimized PID parameters meet the engineering index requirements for stable operation of the closed-loop control system.
[0057] The beneficial effects of the present invention are:
[0058] 1. The PID parameters calculated by the design method proposed in the present invention can be directly applied to the debugging and operation of the closed-loop control system of a nuclear power plant, while solving the availability problem of PID parameters in the design documents of a nuclear power plant and providing an effective theoretical calculation basis.
[0059] 2. The design method proposed in the present invention not only obtains the model parameters of the controlled object of the nuclear power plant by using the simulation data of the nuclear power plant design verification platform, but also optimizes the PID parameters of the closed-loop control system, thereby improving the availability of PID parameters and reducing the debugging workload of on-site personnel and the risk of system operation instability.
[0060] 3. The invention of this patent can not only be used to optimize the PID parameters of standard closed-loop control systems, but also can be used to optimize the PID parameters of other types of closed-loop systems such as cascade double closed-loop control and closed-loop control with feedforward. In addition, the identification data is not limited to the simulation data of the nuclear power plant design verification platform. BRIEF DESCRIPTION OF THE DRAWINGS
[0061] Figure 1 It is a flow chart of a design method for optimal PID parameters of a nuclear power plant closed-loop control system described in a specific embodiment of the present invention;
[0062] Figure 2 It is a schematic diagram of a block diagram of a standard closed-loop control system described in a specific implementation manner of the present invention. DETAILED DESCRIPTION
[0063] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.
[0064] like Figure 1 As shown, the present invention provides a design method for optimal PID parameters of a nuclear power plant closed-loop control system, which is used to obtain the optimal PID parameters of a closed-loop control system that meets engineering index requirements, and includes the following steps:
[0065] Step S1, determine the closed-loop control system that needs to optimize PID parameters, draw a closed-loop control system block diagram (i.e., system principle block diagram) of the closed-loop control system, and determine the controlled object;
[0066] Step S2, performing nuclear power plant simulation on a nuclear power plant design verification platform and operating a closed-loop control system;
[0067] Step S3, determining the model structure and model parameters of the controlled object, using a genetic algorithm to identify the controlled object, obtaining the optimal identified model parameters, and obtaining the final controlled object model;
[0068] Step S4, using genetic algorithm to optimize PID parameters and obtain the optimal PID parameters;
[0069] Step S5, using mathematical software to substitute the controlled object model and the optimized PID parameters into the closed-loop control system block diagram for simulation until the optimized PID parameters meet the engineering index requirements for stable operation of the closed-loop control system;
[0070] Step S6, writing the optimized PID parameters that meet the engineering index requirements for stable operation of the closed-loop control system into the engineering design file.
[0071] In step S1, a closed-loop control system whose PID parameters need to be optimized is determined, and a closed-loop control system block diagram is drawn according to process operating conditions, the controlled object is determined, and the controlled object control input u and the system response output y are clarified. The closed-loop control system block diagram includes a PID controller, an actuator, and a first controlled object. The PID controller, the actuator, and the first controlled object constitute a closed-loop structure model.
[0072] In step S2, a closed-loop control system is simulated and operated on a nuclear power plant design verification platform. When the closed-loop control system is in a steady-state operation stage, an excitation signal is stimulated. The excitation signal can be stimulated at the control input end of the controlled object, can be stimulated by changing the constant value mode of the closed-loop control system, and can also be stimulated by external disturbances. After the excitation signal is triggered, the closed-loop control system gradually transitions from the initial steady state to the new steady state, and records the process change values of the controlled object control input u and the system response output y.
[0073] In step S3, the model structure of the controlled object is determined according to the mechanism characteristics of the controlled object and the form of the system response output y, the model parameters that need to be identified are determined according to the model structure of the controlled object, and the controlled object is identified using a genetic algorithm; when the identification result meets the first evaluation index, the optimal identified model parameters are obtained; the first evaluation index refers to the minimum deviation between the system response output of the controlled object and the response output of the identification system, and the output in the system response output of the controlled object and the response output of the identification system refers to the output signal generated when the input signal excites the controlled object system or the identification system, and the output signal changes over time as an output curve.
[0074] In step S3, the model structure of the controlled object includes the following:
[0075] Expanded form
[0076] Zero-pole form
[0077] Multi-capacity inertia link
[0078] First-order system plus pure delay
[0079] Other forms
[0080] G(s) is the transfer function of the controlled object model
[0081] K is the proportionality factor
[0082] T1…n is the time coefficient
[0083] e -τs It is a delay phase
[0084] n is the transfer function order
[0085] According to the model structure of the controlled object, the model parameters that need to be identified (that is, the identification parameters, including KT 1 T 2 …T n t 0 t 1 …t n-1 τ), use genetic algorithm to identify the controlled object; determine the first evaluation index of genetic algorithm by formula (6): the response output sequence y(k) of the controlled object is compared with the response output sequence y of the identification model * The square sum of the errors of (k) is used as the objective function Q, so that the objective function Q reaches the global minimum; the objective function is selected to determine the fitness function as Q min ;
[0086]
[0087] y(k) is the output sequence of the controlled object obtained by the nuclear power plant design verification platform;
[0088] y * (k) is the output sequence of the controlled object identified by the genetic algorithm;
[0089] C 1 is a constant coefficient to ensure that Q min non-negative;
[0090] is the objective function, which ensures the minimum error of the identification result;
[0091] cov(abs(y(k)-y * (k)) is to suppress noise and ensure that the noise variance is minimized.
[0092] a 1 、b 1 is the error magnification factor to improve the recognition accuracy. 1 、b 1 , so that the identification result quickly reaches the global minimum;
[0093] Set the crossover factor between 0.4-0.9, the mutation factor between 0.01-0.1, and the population size between 30-60. min When it is the smallest, it indicates that the model parameters (KT 1 T 2 …T nt 0 t 1 …t n-1 τ) is optimal; the identified optimal model parameters (KT 1 T 2 …T n t 0 t 1 …t n-1 τ) is substituted into the corresponding controlled object model structure, that is, the final controlled object model is obtained, and the controlled object model includes the controlled object model structure.
[0094] In step S4, PID parameters are optimized using a genetic algorithm through a PID optimization program, the PID parameters found by the genetic algorithm are combined with the controlled object model structure identified by the controlled object, and a closed-loop control system model is calculated; the PID optimization program automatically stimulates a step response for the closed-loop control system; when the output curve of the closed-loop control system meets the second evaluation index, the optimized PID parameters are obtained; the second evaluation index refers to a control index of the quality of the response output curve of the closed-loop control system (the PID optimization program is a program written for PID optimization).
[0095] In step S4, when the genetic algorithm is used to optimize the PID parameters, the closed-loop control system is first simplified into an open-loop system with input r and output y by using the basic simplification rule of the signal flow, where r is the closed-loop fixed value input and y is the system response output y; r is subjected to a unit step, and the step response output curve y of the closed-loop control system is subjected to a second evaluation index: the closed-loop fixed value step input sequence u(k) is compared with the closed-loop calculation model step response output sequence y * The square sum of the errors of (k) is used as the objective function J. When the step response output curve y does not meet the quality requirements, the penalty factor J' is applied to the objective function J to punish it, and finally the objective function J reaches the global minimum. The objective function is selected to determine the fitness function as J min ; The step response output curve y is obtained by the system response output y changing with time (that is, the step response output curve y is an output curve obtained by the output signal of the system response output y changing with time);
[0096]
[0097] u(k) is the closed-loop constant-value step input sequence;
[0098] y * (k) is the step response output sequence of the closed-loop calculation model;
[0099] C 2 is a constant coefficient to ensure that J min non-negative;
[0100] is the objective function, which ensures the minimum error of the identification result;
[0101] cov(abs(u(k)-y * (k)) is to suppress noise and ensure the minimum noise variance;
[0102] a 2 、b 2 is the error magnification factor to improve the recognition accuracy. 2 、b 2 , so that the optimization result can quickly reach the global minimum;
[0103] J' is the penalty factor, which penalizes the step response output curve y that does not meet the engineering quality index;
[0104] The step response output curve y should meet the following quality index requirements as shown in formula (8):
[0105]
[0106] M p is the overshoot of the step response output curve y,
[0107] φ is the attenuation rate of the step response output curve y,
[0108] u pid is the output capability of the PID controller,
[0109] t s-pid It is the average time for the PID controller to change 100% opening;
[0110] The penalty factor J' can be adjusted and changed according to engineering requirements;
[0111] When the fitness function J min When it is the smallest, it means that the PID parameters obtained by the genetic algorithm are the optimal PID parameters. Substitute the optimal PID parameters into the PID controller in the closed-loop control system block diagram ( Figure 2 In the mathematical model of the PID controller), the final closed-loop structure model is obtained ( Figure 2 The PID controller, the actuator, and the first controlled object constitute a closed-loop structure model; in engineering applications, the actuator structure and the first controlled object are usually identified as a whole, that is, the second controlled object).
[0112] In step S5, mathematical software (MATLAB) is used for simulation verification, a closed-loop control system block diagram is built, and the controlled object model and the optimized PID parameters are substituted into the closed-loop control system block diagram for simulation. If the step response output curve characteristics of the closed-loop control system do not meet the engineering index requirements, the second evaluation index can be adjusted according to the engineering requirements. After the adjustment, the genetic algorithm is used again to optimize the PID parameters until the optimized PID parameters meet the engineering index requirements for stable operation of the closed-loop control system.
[0113] Finally, in step S6, after the above steps are completed, the optimal PID parameters can be finally written into the engineering design file. The parameters have been verified by simulation and can be directly applied to the commissioning and operation of the nuclear power plant.
[0114] The method described in the present invention is not limited to the embodiments described in the specific implementation manner. Those skilled in the art may derive other implementation manners based on the technical solution of the present invention, which also fall within the technical innovation scope of the present invention.
Claims
1. A design method for optimal PID parameters of a closed-loop control system of a nuclear power plant, used to obtain the optimal PID parameters of a closed-loop control system that meets engineering index requirements, comprising the following steps: Step S1, determining a closed-loop control system whose PID parameters need to be optimized, drawing a closed-loop control system block diagram of the closed-loop control system, and determining a controlled object; Step S2, performing a nuclear power plant simulation on a nuclear power plant design verification platform and running the closed-loop control system; Step S3, determining the model structure and model parameters of the controlled object, identifying the controlled object using a genetic algorithm, obtaining the optimally identified model parameters, and acquiring the final controlled object model; Step S4, using the genetic algorithm to optimize PID parameters and obtain the optimal PID parameters; Step S5, using mathematical software to substitute the controlled object model and the optimized PID parameters into the closed-loop control system block diagram for simulation until the optimized PID parameters meet the engineering index requirements for stable operation of the closed-loop control system; Step S6, writing the optimized PID parameters that meet the engineering index requirements for stable operation of the closed-loop control system into an engineering design document; In the step S1, a closed-loop control system whose PID parameters need to be optimized is determined, and a block diagram of the closed-loop control system is drawn according to process operating conditions, a controlled object is determined, and a control input u of the controlled object and a system response output y are clarified. The closed-loop control system block diagram includes a PID controller, an actuator and a first controlled object, and the PID controller, the actuator and the first controlled object form a closed-loop structure model; In the step S2, the closed-loop control system is simulated and operated in a nuclear power plant on a nuclear power plant design verification platform. When the closed-loop control system is in a steady-state operation stage, an excitation signal is stimulated. The excitation signal can be stimulated at the control input end of the controlled object, can also be stimulated by changing the fixed value mode of the closed-loop control system, and can also be stimulated by using external disturbances. After the excitation signal is triggered, the closed-loop control system gradually transitions from the initial steady state to the new steady state, and records the process change value of the controlled object control input u and the system response output y. In the step S3, the model structure of the controlled object is determined according to the mechanism characteristics of the controlled object and the form of the system response output y, the model parameters to be identified are determined according to the model structure of the controlled object, and the controlled object is identified using a genetic algorithm; when the identification result meets the first evaluation index, the optimally identified model parameters are obtained; the first evaluation index refers to the minimum deviation between the system response output of the controlled object and the response output of the identification system, the output of the system response output of the controlled object and the response output of the identification system refers to the output signal generated when the input signal excites the controlled object system or the identification system, and the output signal changes over time as an output curve; In step S3, the controlled object model structure includes the following: Expanded form Zero-pole form Multi-capacity inertia link First-order system plus pure delay Other forms G(s) is the transfer function of the controlled object model K is the proportionality factor T 1…n is the time coefficient e -τs It is a delay phase n is the transfer function order Determine the model parameters (KT) that need to be identified according to the model structure of the controlled object 1 T 2 …T n t 0 t 1 …t n-1 τ), using the genetic algorithm to identify the controlled object; determining the first evaluation index of the genetic algorithm by formula (6): comparing the controlled object response output sequence y(k) with the identification model response output sequence y * The square sum of the errors of (k) is used as the objective function Q, so that the objective function Q reaches the global minimum; the objective function is selected to determine the fitness function as Q min ; y(k) is the output sequence of the controlled object obtained by the nuclear power plant design verification platform; y * (k) is the output sequence of the controlled object identified by the genetic algorithm; C 1 is a constant coefficient to ensure that Q min non-negative; is the objective function, which ensures the minimum error of the identification result; cov(abs(y(k)-y * (k)) is to suppress noise and ensure the minimum noise variance; a 1 , b 1 is the error magnification factor to improve the recognition accuracy. 1 , b 1 , so that the identification result quickly reaches the global minimum; Set the crossover factor between 0.4-0.9, the mutation factor between 0.01-0.1, and the population size between 30-60. min When it is the smallest, it indicates that the model parameter (KT 1 T 2 …T n t 0 t 1 …t n-1 τ) is optimal; the identified optimal model parameter (KT 1 T 2 …T n t 0 t 1 …t n-1 τ) is substituted into the corresponding controlled object model structure, that is, the final controlled object model is obtained, and the controlled object model includes the controlled object model structure.
2. The design method of optimal PID parameters of a closed-loop control system of a nuclear power plant as claimed in claim 1, Its characteristics are: In the step S4, the PID parameters are optimized by the genetic algorithm through a PID optimization program, the PID parameters found by the genetic algorithm are combined with the controlled object model structure identified by the controlled object, and a closed-loop control system model is calculated; the PID optimization program automatically stimulates a step response for the closed-loop control system; when the output curve of the closed-loop control system meets the second evaluation index, the optimized PID parameters are obtained; the second evaluation index refers to a control index of the quality of the response output curve of the closed-loop control system.
3. The design method of optimal PID parameters of a closed-loop control system of a nuclear power plant as claimed in claim 2, Its characteristics are: In step S4, when the genetic algorithm is used to optimize the PID parameters, the closed-loop control system is first simplified into an open-loop system with an input of r and an output of y by using the basic simplification rule of the signal flow, wherein r is a closed-loop fixed-value input quantity and y is the system response output y; a unit step is performed on r, and the step response output curve y of the closed-loop control system is subjected to the second evaluation index: the closed-loop fixed-value step input sequence u(k) is compared with the closed-loop calculation model step response output sequence y * The square sum of the errors of (k) is used as the objective function J. When the step response output curve y does not meet the quality requirements, the objective function J is penalized by applying a penalty factor J', and finally the objective function J is made to reach the global minimum. The objective function is selected to determine the fitness function J min ; The step response output curve y is obtained by the change of the system response output y over time; u(k) is the closed-loop constant-value step input sequence; y * (k) is the step response output sequence of the closed-loop calculation model; C 2 is a constant coefficient to ensure that J min non-negative; is the objective function, which ensures the minimum error of the identification result; cov(abs(u(k)-y * (k)) is to suppress noise and ensure the minimum noise variance; a 2 , b 2 is the error magnification factor to improve the recognition accuracy. 2 , b 2 , so that the optimization result can quickly reach the global minimum; J' is a penalty factor, which penalizes the step response output curve y that does not meet the engineering quality index; The step response output curve y should meet the following quality index requirements as shown in formula (8): M p is the overshoot of the step response output curve y, φ is the attenuation rate of the step response output curve y, u pid is the output capability of the PID controller, t s-pid It is the average time for the PID controller to change 100% opening; When the fitness function J min When it is the smallest, it indicates that the PID parameters obtained by the genetic algorithm are the optimal PID parameters. The optimal PID parameters are substituted into the PID controller in the closed-loop control system block diagram to obtain the final closed-loop structure model.
4. The design method of optimal PID parameters of a closed-loop control system of a nuclear power plant as claimed in claim 2, Its characteristics are: In the step S5, mathematical software is used for simulation verification, the closed-loop control system block diagram is built, the controlled object model and the optimized PID parameters are substituted into the closed-loop control system block diagram for simulation, if the step response output curve characteristics of the closed-loop control system do not meet the engineering index requirements, the second evaluation index is adjusted according to the engineering requirements, and after the adjustment, the genetic algorithm is used again to optimize the PID parameters until the optimized PID parameters meet the engineering index requirements for stable operation of the closed-loop control system.
