Application of Fuzzy Controllers for Power Control of the VVER-1200 Nuclear Reactor

Abstract

This article deals with fuzzy and fuzzy PI controllers for the automatic power control system of a linearized mathematical model of the VVER-1200 nuclear reactor. The automatic power control system includes a mathematical model of a step-by-step electromagnetic drive (SED), a model of in-core and ex-core neutron flux sensors, and a refined mathematical model of the group 12 of control and protection system of control rods (CPS CR) obtained by approximating experimental data using the Levenberg–Marquardt algorithm for the nonlinear least squares problem. Based on the state space model, ten transfer function matrices of the nuclear reactor corresponding to a power range of from 10 to 100% of the nominal power were determined, and ten classical PI controllers were designed to ensure stability margins of at least 60° in phase and at least 10 decibels in amplitude for each power level of the plant. Simulation results show a significant advantage of the developed fuzzy controllers both in the steady-state power maintenance mode considering noise in the neutron flux sensor channel and in the load-following mode at different power levels.