Model-based control of a research nuclear reactor

Abstract

The purpose of this paper is to generate a controller to raise the reactor neutronic power from a lower level to a higher level and stabilize it a that level. The dynamic response of research reactors was studied during and after the power transient process. An algorithm for the control process was generated, based on a previously published controller. The inverse-kinetics model-based controller was tested by simulation in order to verify successful completion of the control task, during and after the power transient process. A Kalman filter system was constructed to estimate the state parameters needed by the controller. System dynamic response under different conditions was obtained through simulations. The effects of control performance and system dynamic response of model mismatch between the reactor system and the filter system and of incorrectness of initializations of the filter system were studied. Selection of a filter system with an appropriate Kalman gain matrix was made after conducting several simulation tests. Comparison was made between experimental results obtained from the University of Arizona TRIGA research reactor and simulation results from the computerized simulation systems.