Fuzzy clustering-based large-scale multimodal multi-objective differential evolution algorithm

作     者：Wu, Lingyu Zhao, Xinchao Ye, Lingjuan Qiao, Zenglin Zuo, Xingquan 

作者机构：Beijing Univ Posts & Telecommun Sch Sci Beijing 100876 Peoples R China Beijing Univ Posts & Telecommun Key Lab Math & Informat Networks Minist Educ Beijing Peoples R China Beijing Univ Posts & Telecommun Sch Comp Sci Beijing 100876 Peoples R China 

出 版 物：《SWARM AND EVOLUTIONARY COMPUTATION》 (Swarm Evol. Comput.)

年 卷 期：2025年第93卷

核心收录：

学科分类：08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：National Natural Science Foundation of China [62476030, 62303066] Beijing Natural Science Foundation Fundamental Research Funds for the Central Unverities [2023RC46] 

主　　题：Multimodal multi-objective optimization Differential evolution Large-scale optimization Fuzzy C-means Evolutionary computation 

摘      要：Traditional multi-modal multi-objective problems often have multiple local optima in the decision space, and each local optimum is of great importance. However, real-world multi-modal problems are often largescale problems, and there are few algorithms specifically designed for large-scale multi-modal multi-objective problems. Some proposed algorithms have been tested only on specific problems and are not applicable to solve other specific problems. Based on this problem, this paper proposes a large-scale multi-modal multi- objective differential evolution algorithm called LMMODE, based on fuzzy clustering. The Fuzzy C-means(FCM) algorithm, suitable for high-dimensional data, is employed to divide the search space into multiple subspaces. The multi-stage optimization approach is then utilized to balance the algorithm s performance in the objective space and decision space through different strategies, thereby solving large-scale multi-modal multi-objective problems. Experimental results demonstrate that, compared to state-of-the-art multi-modal and large-scale algorithms, LMMODE is competitive in solving large-scale multi-modal problems.