The Research of PWR Steam Generator with Water Level Control Based on RBF Neural Network

Abstract

In the steam generator with water level control system of nuclear power plant, there are various uncertainties in the controlled devices. Any actual system had certain nonlinear. Because of the steam generator is very important equipment in nuclear power plants, water level control plays a decisive role for the safe operation of nuclear power plant and it required stable operation and fast response of the whole system. For the highly complex, non-linear system , the traditional cascade PID control had been used cannot obtain satisfactory control effect, this paper try to use RBF neural network to optimize the PID parameters. The simulation experiments show that: the rbf neural network optimized controller made the control system’s robustness and control quality superior than the traditional PID controller, and described the method can be applied more widely.