Brain Storm Optimization Algorithm Based on Competition Mechanism.

Fenqiang Li,Hu Peng,Changshou Deng,Yi Zeng,Yucheng Tan,Wenjun Wang
DOI: https://doi.org/10.1109/cec.2019.8790330
2019-01-01
Abstract:Brain storm optimization algorithm (BSO) is a new but excellent algorithm who based on brainstorming process. However, BSO has some disadvantages like premature convergence and high time complexity. In order to solve these problems, a new BSO based on competition mechanism (BSOCM) is proposed in this paper. BSOCM replaces the k-means cluster by sorting the population, and adopts the mutation strategy of competition mechanism to stimulate the population to generate more individuals. In this way, BSOCM reduces time complexity. At the same time, the strategy of competition mechanism is added to ensure the optimization ability of the algorithm, and the precocity of the algorithm is reduced. The experimental results showed that BSOCM could avoid premature local optimization. The benchmark function of CEC 2013 was used in a comprehensive experiment, the test results of BSO, DE, MBSO, PSO and BSOCM were compared respectively. Relevant experiments in this paper prove that BSOCM has good search efficiency.
What problem does this paper attempt to address?