文献类型：期刊文献

英文题名：Research on Multi-Objective Evolutionary Algorithms Based on Large-Scale Decision Variable Analysis

作者：Li, Jianing[1];Xu, Sijia[1];Zheng, Jiaming[1];Jiang, Guoqing[2];Ding, Weichao[1,2]

机构：[1]East China Univ Sci & Technol, Sch Informat Sci & Engn, Shanghai 200237, Peoples R China;[2]Minist Publ Secur, Key Lab Informat Network Secur, Shanghai 200031, Peoples R China

年份：2024

卷号：14

期号：22

外文期刊名：APPLIED SCIENCES-BASEL

收录：;EI(收录号：20244917456914);WOS:【SCI-EXPANDED(收录号:WOS:001366912300001)】；

基金：This work was sponsored by the National Natural Science Foundation of China (No. 62403201); the Shanghai Pilot Program for Basic Research, China, 22TQ140010016; the Nature Science Foundation of Shanghai, China, 23ZR1414900, 22ZR1416500; and Key Lab of Information Network Security, Ministry of Public Security, China, C23600-03.

语种：英文

外文关键词：dual-population cooperative evolution; decision variable characteristic analysis; large-scale optimization; high-dimensional many-objective; container resource scheduling

摘要：Large-scale high-dimensional many-objective optimization problems (LaMaOPs) are prevalent in fields such as autonomous driving, cloud resource scheduling, and smart grids. LaMaOPs involve a large number of decision variables and multiple conflicting objectives that need to be optimized simultaneously. The challenges posed by the curse of dimensionality due to the vast number of decision variables, and the conflict between convergence and diversity caused by the numerous objective variables, make traditional optimization methods inadequate. To address these issues, this paper proposes a two-population cooperative evolutionary algorithm based on large-scale decision variable analysis (DVA-TPCEA). This algorithm integrates quantitative analysis methods for decision variables to deeply examine their impact on each objective and introduces a contribution-based objective detection method. Additionally, a dual-population cooperative evolution mechanism is employed, with targeted optimization strategies designed for convergence and diversity populations, achieving synergistic and complementary optimization between the two populations. To validate the algorithm's effectiveness in practical applications, a large-scale container resource scheduling strategy based on the DVA-TPCEA algorithm is also proposed. The experimental results indicate that the proposed algorithm demonstrates significant advantages in both general datasets DTLZ, WFG, and LSMOP, and practical models.