Pressurized Water reactor power level control: A nonlinear generalized predictive control with extended Kalman filter method

Abstract

Reactor power level control is an effective way to achieve load tracking of Pressurized Water Reactor (PWR) in a nuclear power station. A novel Nonlinear Generalized Predictive Control with Extended Kalman Filter (NGPC + EKF) is proposed to solve the problem that discrete predictive model mismatch in noisy environment. In this paper, an NGPC controller is developed to realize the reactor load tracking, and an EKF is used to estimate reactor states and suppress noise. Finally, the control methods of PID, MPC, NGPC and NGPC + EKF are compared by two simulation experiments, load tracking experiment and step response experiment. The load tracking experiment results show that NGPC + EKF method obtains better noise suppression ability and tracking effect. In the step response experiment, the proposed NGPC + EKF scheme is also proved to have better step response performance than others.