Enhancing the local search ability of the brain storm optimization algorithm by covariance matrix adaptation

Abstract

Recently, the Brain Storm Optimization (BSO) algorithm has attracted many researchers and practitioners attention from the evolutionary computation community. However, like many other population based algorithms, BSO shows good performance at global exploration but not good enough at local exploitation. To alleviate this issue, in this chapter, the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) is utilized in the Global-best BSO (GBSO), with the aim to combine the exploration ability of BSO and local ability of CMA-ES and to design an improved version of BSO. The performance of the proposed algorithm is tested by solving 28 classical optimization problems and the proposed algorithm is shown to perform better than GBSO.