Improving Dual-Population Differential Evolution Based on Hierarchical Mutation and Selection Strategy

• DOI: 10.3390/electronics13010062
• Publication Date: 2023-12-22T06:19:42Z
• Authors: Yawei Huang; Xuezhong Qian; Wei Song
• Keywords: Differential Evolution; Local optimum; Tournament selection

The dual-population differential evolution (DDE) algorithm is an optimization technique that simultaneously maintains two populations to balance global and local search. It has been demonstrated outperform single-population algorithms. However, existing improvements algorithms often overlook the importance of selecting appropriate mutation selection operators enhance performance. In this paper, we propose a (DPDE) based on hierarchical strategy. We divided population into elite normal subpopulations fitness values. Information exchange between was facilitated through strategy, promoting balanced exploration–exploitation trade-off in algorithm. Additionally, paper presents new strategy aimed at improving population’s capacity avoid optima. achieves by accepting discarded trial vectors differently compared previous methods. expect newly introduced strategies will work synergy, effectively harnessing their potential algorithm’s Extensive experiments were conducted CEC 2017 2011 test sets. results showed DPDE offers competitive performance, comparable six state-of-the-art