A hybrid multi-objective optimization method for nuclear essential service water system design

Abstract Evolutionary algorithms have proven to be very successful in solving multi-objective optimization problems (MOPs). However, their performance often deteriorates when constraints are introduced. This paper proposes a hybrid heuristic intelligence (HHI) approach remedy this issue. First, classification operator is designed sort individuals by relative constraint values, which enlarges the size of feasible solutions early process. Then, combining with non-dominated sorting method, we define new selection increase probability selecting who low values but high objective values. Finally, use optimized operations simultaneously enhance convergence and diversity performing different operators certain domain. The proposed method achieves state-of-the-art on test functions ZDT1, Binh2, OSY, remarkable decrease 24% inverted generational distance (IGD) OSY. For further applications, our also tested an engineering model established for design nuclear essential service water system (SEC). final result shows that without taking hundreds hours, solution HHI has met all reduced total cost 8.9%© YEAR Authors. Published Elsevier Ltd.