Optimal membership function based fuzzy proportional–integral controller for power control of molten salt breeder reactor core

Abstract

The aim of this paper is to design an optimal membership function (MF)-based fuzzy PI (proportional integral) controller to control the core power of a nuclear reactor. The molten salt breeder reactor (MSBR) is extensively utilized for studying primary power management of nuclear reactors. Many challenges are associated with such systems, including mathematical modelling errors, parametric uncertainties, and external disturbances. Moreover, the liquid fuel used in MSBR systems makes it more challenging to design a suitable controller for the core relative power control of such systems. The conventional PI controller may not perform efficiently under such uncertainties in the MSBR system. Different types of perturbations can be handled by a fuzzy controller, if the coordination among fuzzy rules and membership functions of fuzzy variables is done accurately. The idea of this work is to optimize the settings (range and scale) of fuzzy MFs in a fuzzy base controller. The fuzzy controller will not be used directly in the system; rather, it will be used to tune a PI controller for the system where proper expert knowledge may not be available. A nonlinear dynamic acceleration coefficients-based class topper optimization (NDAC-CTO) algorithm is also developed to optimize the fuzzy membership functions. For load tracking with step disruption, the power control of the MSBR core is investigated. With the proposed controller, a significant improvement of 70 to 80 % in the settling time of the power profile of the MSBR system is achieved in comparison with the existing results.