Multivariant Optimization Algorithm with Bimodal-Gauss

In multimodal problems, there is a trade-off between exploration and exploitation. Exploration contributes to move quickly toward the area where better solutions existed but is not beneficial for improving the quality of intermediate solution. Exploitation do well in refine the intermediate solution but increase the risk of being trapped into local optimum. Considering the trade-off and advantage of exploration and exploitation, a local search strategy based on bimodal-gauss was embedded into multivariant optimization algorithm by increasing the probability of locating global optima in solving multimodal optimization problems. The performances of the proposed method were compared with that of other multimodal optimization algorithms based on benchmark functions and the experimental results show the superiority of the proposed method. Convergence process of each subgroup was analyzed based on convergence curve.