Data-driven-based methodology to improve performance of reactor power regulation system in small pressurized water reactor

Abstract Frequent and wide-range load changes in small pressurized water reactors applied for nuclear-powered devices are necessary under complex and harsh conditions. However, the traditional power regulation method does not consider the stroke of the control rod, which seriously affects the service life of the control rod drive mechanism. This paper proposes a data-driven reactor power regulation system (RPRS) based on the hybrid feedforward and feedback control theory. Different adjustment regions are set in the controller to adjust the proportion of the feedforward and feedback signals dynamically. The polynomial fitting method and particle swarm algorithm are applied to optimize the size of the different adjustment regions. Typical load change conditions are applied to test and verify that the data-driven RPRS can efficiently shorten the overshoot of the control rod position and improve the performance of the power regulation system. The robustness of the data-driven RPRS is also demonstrated by simulation.