期刊
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
卷 19, 期 1, 页码 50-73出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TEVC.2014.2301794
关键词
Diversity; local search; multiobjective optimization; nondominated sorting; test problems
In this paper, a new multiobjective optimization framework based on nondominated sorting and local search (NSLS) is introduced. The NSLS is based on iterations. At each iteration, given a population P, a simple local search method is used to get a better population P', and then the nondominated sorting is adopted on P. P' to obtain a new population for the next iteration. Furthermore, the farthest-candidate approach is combined with the nondominated sorting to choose the new population for improving the diversity. Additionally, another version of NSLS (NSLS-C) is used for comparison, which replaces the farthest-candidate method with the crowded comparison mechanism presented in the nondominated sorting genetic algorithm II (NSGA-II). The proposed method (NSLS) is compared with NSLS-C and the other three classic algorithms: NSGA-II, MOEA/D-DE, and MODEA on a set of seventeen bi-objective and three tri-objective test problems. The experimental results indicate that the proposed NSLS is able to find a better spread of solutions and a better convergence to the true Pareto-optimal front compared to the other four algorithms. Furthermore, the sensitivity of NSLS is also experimentally investigated in this paper.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据