A Neighborhood-Based Speciation Brain Storm Optimization with Evolution Strategy for Multimodal Optimization
Honglin Jin, Shi Cheng, Xue-ping Wang, Yue Liu, Yuyuan Shan, Hao Ran, Hui Lu · 2023
Finding multiple optimal solutions is challenging for solving multimodal optimization problems (MMOPs). In this paper, a neighborhood-based speciation brain storm optimization with evolution strategy (NS-BSO-ES) is proposed to solve MMOPs, which combines the advantages of better exploration of the neighborhood-based speciation brain storm optimization (NS-BSO) and more robust exploitation of the evolution strategy with covariance matrix adaptation (CMA-ES). In NS-BSO-ES, NS- BSO is used to generate candidate solutions to maintain the diversity of the population, CMA-ES is adopted to enhance the local search ability and locate optimal solutions accurately, and the archive is used to store inferior solutions to fully utilize the valuable information contained in these solutions as potential directions towards the optimal solution. To test the performance of NS-BSO-ES for solving MMOPs, compared with related algorithms on the 20 benchmark MMOPs in CEC-2013 Functions. Experimental results indicate NS-BSO-ES outperforms the other compared algorithms on most tested benchmark functions.