关键词: 多目标优化, 混合选择, 进化算法, 工程设计

Abstract: To solve the complex multi-objective optimization design problem in actual engineering problem,a Multi-Objective Evolutionary Algorithm based on Hybrid Selection (HSMEA) was proposed.A dynamic strategy which combined angle and distance was implemented to assist the algorithm to cluster the individuals.Thereafter,the hybrid selection mechanism that included two different principles was conducted to further select the solutions.Thus,solutions were obtained with the better balance of convergence and diversity.The proposed method was compared with other four state-of-the-art multi-objective evolutionary algorithms on a number of test problems.The experimental results showed that the competitiveness and effectiveness of the proposed algorithm in addressing the multi-objective problems.Additionally,HSMEA was implemented in the engineering design optimization cases,which verified the well performance and the potential of the proposed algorithm in real-world optical problems.

Key words: multi-objective optimization, hybrid selection, evolutionary algorithm, engineering design