Data-driven Based Performance Optimization Method For Steam Generator Liquid Level Control System

Steam generator is an essential component of a nuclear power plant.The regular operation of the steam generator is of great significance to the safety of the nuclear power plant.Among them,the liquid level of the steam generator is a key control parameter in the operation process;Too high or low liquid level will affect the safety of nuclear power plant operation.Therefore,the level control of the steam generator of the nuclear power plant is critical to the operational safety of the nuclear power plant.In steam generator liquid level control,a controller is used to control the liquid level.Most controllers require to adjust their parameters.The selection of parameters affects the performance of the liquid level control system.Therefore,obtaining a stable control performance requires determining its optimal control parameters.Traditional control parameter tuning can be divided into three categories,trial and error method,empirical formula method,and model-based optimization method.The empirical formula method is mainly used for PID control parameter tuning.Still,the traditional tuning method has a high cost,low efficiency,and the defect of relying heavily on experience.To solve the above problems,improve parameter tuning efficiency and reduce the optimization cost,this paper combines the data-driven ideology.It proposes a data-driven steam generator liquid level control system performance optimization method.Based on the above optimization method,a systematic performance optimization architecture of liquid level control system is formed.Firstly,on the basis of analyzing the characteristics of the steam generator of nuclear power plants,establish a mechanism model of the steam generator liquid level,and implement the cascade three-impulse liquid level control scheme based on this basis,and verify the above model and control system design,To lay the foundation for the research of control system performance optimization methods;Secondly,on the basis of refining the performance optimization problem of the steam generator liquid level control system,a systematic data-driven optimization framework are proposed;an iteration termination control mechanism based on a rolling iteration termination factor is designed,Realize the intelligent and autonomous control of the optimization process;Thirdly,summarize the thought,guidance and fusion mechanism of the data-driven optimization method,and on this basis,propose three Simultaneous Perturbation Stochastic Approximation(SPSA)methods based on different guidance mechanisms to form three data-driven steam generation methods.The performance optimization method of the liquid level control system of the device,and the optimization performance of the three methods is tested and analyzed;Fourthly,design and implement a steam generator liquid level MPC control system,form a data-driven steam generator MPC control system performance optimization framework,apply three data-driven optimization methods to the performance optimization of the MPC control system and test its performance analysis.At last,through summarizing and analyzing,the prospect of the optimization direction of the performance of the steam generator liquid level control system is put forward.