An optima-identified framework with brain storm optimization for multimodal optimization problems

Locating multiple optima and maintaining these identified solutions are two crucial issues in solving multimodal optimization problems (MMOPs). To address these two challenges, an optima-identified framework (OIF) combined with brain storm optimization (BSO) algorithm is proposed in this paper, which can identify global optimal solutions found during the search process, and maintain these optima until the end of the run. First, a max-fitness clustering method (MCM) is applied to form different clusters and each cluster center is likely to become an extreme point. Then, a modified disruption strategy (MDS) is devised to distinguish and identify potential optima among these cluster centers. Finally, we introduce two kinds of redistribution strategies (RS) to make the most of the individuals in those clusters whose cluster centers have been identified as global optima. We validate the effectiveness of the OIF and compare the proposed OIF-BSO algorithm with other state-of-the-art multimodal optimization algorithms. The results indicate that our framework is feasible to maintain and utilize cluster centers over the course of search, and the proposed algorithm can outperform other algorithms.

Automated summary

In this paper, an optima-identified framework (OIF) combined with brain storm optimization (BSO) algorithm is proposed to address the crucial issues of locating multiple optima and maintaining these identified solutions in multimodal optimization problems (MMOPs). The framework effectively identifies and utilizes global optimal solutions, distinguishes potential optima through clustering and disruption strategies, and maximizes the use of individuals in clusters where cluster centers are identified as global optima.