A species-based multi-objective genetic algorithm for multi-objective optimization problems
Fuquan Sun, Wang Hongfeng, Fuqiang Lu · 2014
In recent years, multi-objective optimization problems (MOOPs) have gained a lot of attentions from the community of evolutionary algorithm since many real-world optimization problems would involve multiple objective functions. In this paper, a species-based multi-objective genetic algorithm (SMOGA) that hybridizes a species method, which was initially designed in GA for multi-modal problems, with the algorithm mechanism of NSGA-II, which was one of well-known MOGAs, is proposed for MOOPs. In order to examine the performance of the proposed algorithm, experiments were carried out to investigate the strength and weakness of SMOGA on a series of test MOOPs in comparison with NSGA-II.