An adaptive multiobjective evolutionary algorithm based on grid subspaces

The successful application of multi-objective evolutionary algorithms (MOEAs) in many kinds of multiobjective problems have attracted considerable attention in recent years. In this paper, an adaptive multi-objective evolutionary algorithm is proposed by incorporating the concepts of the grid system (denoted as AGMOEA). Based on grid, the objective space is divided into subspaces. Based on the quality and dominance relationship between subspaces, the evolutionary opportunities are dynamically allocated to different subspaces with an adaptive selection strategy. To improve the evolutionary efficiency, the evolutionary scheme and an external archive mechanism considering representative individuals are proposed. The experimental results on 21 benchmark problems demonstrate that the proposed algorithm is competitive or superior to the rival algorithms.

Automated summary

The paper proposes an adaptive multi-objective evolutionary algorithm AGMOEA, which divides the objective space into subspaces based on the grid system, dynamically allocates evolutionary opportunities to different subspaces based on relationships between them. Experimental results demonstrate the algorithm's strong competitiveness.