A non-dominated discrete differential evolution for fuel loading pattern optimization of a nuclear research reactor

Abstract This paper presents a detailed description of a new variant of differential evolution for nuclear reactor refueling optimization problem. This variant combines the elitism strategy with a discrete differential evolution. The elitism strategy allows non-dominated solutions found during the search and stored in the archive to participate in the differential evolution operation. The population size is the same as the archive size, and the number of non-dominated solutions participating in the search at a particular generation is controlled by a specific probability. The proposed method is successfully applied to a nuclear research reactor for its first refueling time to search for optimal loading patterns that both maximize the effective multiplication k eff and minimize the power peaking factor PPF of the reactor. The optimal loading patterns can significantly improve the operational time and safety of the reactor compared to the loading pattern used in practice.