Chaotic-NSGA-II: An effective algorithm to solve multi-objective optimization problems

Abstract

This paper presents a new approach to handle multi-objective optimization problems (MOP) by incorporating logistic mapping function into the process of NSGA-II. NSGA-II is a well-known evolutionary algorithm for optimization, it is famous for its small computational complexity and simpleness, its ability to maintain a good spread of solutions makes it converge better in the obtained non-dominated front than PAES and SPEA. But it may lack of diversity, so we introduce chaos into this NSGA-II, aiming to add chaos to the solutions generated by the genetic process. Chaos optimization algorithm (COA) was proposed that can solve complex function optimization and has a high efficiency of calculation. Through the comparison of Chaotic-NSGA-II and NSGA-II on six test problems, we can see that this algorithm has a good performance on searching the global optimization.