A Multi-modal Multi-objective Evolutionary Algorithm Based on Multi-criteria Grouping

Multi-modal multi-objective optimization problems (MMOPs) are a distinct subset of multi-objective problems, where one objective value often corresponds to multiple solutions. Consequently, achieving diversity within the population is crucial in both the decision and objective spaces when tackling MMOPs. In this paper, we introduce a novel multi-modal multi-objective evolutionary algorithm based on a multi-criteria grouping technique. Our proposed method involves dividing the population into several sub-groups in both the decision space and the objective space, followed by the selection of a demonstrator for each sub-group. Subsequently, different mutation strategies are applied in different stages to update the position of each solution. The experimental results on 22 test functions show that our algorithm outperforms five state-of-the-art algorithms for solving multi-modal multi-objective problems.