Brain Storm Optimization Algorithm with Re-initialized Ideas and Adaptive Step Size

Abstract

Brain Storm Optimization (BSO) is a recently developed population-based algorithm to mimic the brainstorming process in humans. It has been successfully applied in the domain of non-linear continuous optimization. In this work, we propose enhancing the performance of BSO by introducing a re-initialization mechanism triggered by the current state of the population. In addition, we also propose to modify the step-size equation in order to take the search space size into consideration. The proposed improved BSO is compared with two of the most recent BSO variants based on the CEC15 benchmarks.