Species-based differential evolution with switching search strategies for multimodal function optimization

In multimodal optimization problems, the objective is not only to find one global optimal solution, but also to find various global optimal solutions. For this purpose, the Species-based Differential Evolution (SDE) has been proposed previously. In this method, the population is divided into multiple subpopulations by using speciation, and each species focuses its search for one optimal solution. By this way, multiple optimal solutions can be discovered simultaneously. However, this algorithm takes a long time for complicated problems to acquire global optima. In this paper, we propose SDE with switching search strategies, which selects global search by the population or local search in each species according to the search situation. The comparison of the conventional SDE and our proposed method is performed on five test functions. The experimental results show that the SDE with switching search strategies outperforms the conventional SDE in a complicated function problem.