文献类型：期刊文献

英文题名：A novel path-based reproduction operator for multi-objective optimization

作者：Song, Wenjiang[1];Du, Wei[1,2];Fan, Chen[1];Zhong, Weimin[1];Qian, Feng[1]

机构：[1]East China Univ Sci & Technol, Minist Educ, Key Lab Adv Control & Optimizat Chem Proc, Shanghai 200237, Peoples R China;[2]Tongji Univ, Shanghai Inst Intelligent Sci & Technol, Shanghai 200092, Peoples R China

年份：2020

卷号：59

外文期刊名：SWARM AND EVOLUTIONARY COMPUTATION

收录：;EI(收录号：20203309050301);WOS:【SCI-EXPANDED(收录号:WOS:000599930700003)】；

基金：This work was supported by National Key R&D Program of China (2016YFB0303401), National Natural Science Foundation of China (61890930-3), International (Regional) Cooperation and Exchange Project (61720106008), and the Fundamental Research Funds for the Central Universities.

语种：英文

外文关键词：Multi-objective optimization; Evolution path; Reproduction operator; Gene-sharing; Parameter self-adaptation

摘要：A large number of multi-objective evolutionary algorithms (MOEAs) have been proposed for the past two decades. However, few papers focus on the study of reproduction operator in MOEAs. In this work, we propose a novel path-based reproduction operator, termed path evolution (PE), to generate potential solutions more effectively for MOEAs. In PE, there is no mating selection, and the calculation of the evolution path is simple. Moreover, a new gene-sharing operation is proposed. The effectiveness of PE is validated by comparing it with three widely used reproduction operators and two state-of-the-art path-based reproduction operators. It is also reported that PE is very flexible to embed into different categories of MOEAs. The empirical results on three widely used test suites demonstrate the superiority, especially faster convergence ability, of PE.