Control of the reactor core power in PWR using optimized PID controller with the real-coded GA

Abstract

Load following is an importance topic in the Nuclear Power Plants (NPPs). One of the conventional and simplest ways is the use of Proportional-Integral-Derivative (PID) controller. The reactor power is simulated based on the point kinetic model. PID gains of a nonlinear time-varying system (a PWR NPP) are optimized and scheduled using real-coded genetic algorithm (GA). To this end, the objective function of the decision variables, include the overshoot, settling time and stabilization time (based on the Lyapunov approach) of the system is minimized. The presented control system track demand power level change within a wide range of time. The simulation results demonstrate good stability of this method and show high performance of the optimized PID gains to adapt any changes in the output power.