A many-objective evolutionary algorithm with diversity-first based environmental selection

Abstract Environmental selection in Pareto-based many-objective evolutionary algorithms generally employ Pareto-dominance relation to first consider the convergence and give higher priority to convergence than diversity. When the many-objective optimization problem has a complicated Pareto front, this selection strategy can easily miss the promising areas and converge into a subregion of the Pareto front. To address this issue, we propose a many-objective evolutionary algorithm with diversity-first based environmental selection. Different from the existing selection strategies, the environmental selection procedure in the proposed algorithm adopts a diversity-first-and-convergence-second principle, which first selects the representative solutions that having better diversity and then considers using the well-converged solutions to replace them in subregions. This selection-replacement strategy can maintain the diversity and make contribution to the convergence. In addition, a selection criterion, termed adaptive angle penalized distance, is designed to judge whether the replacement is implemented or not. The proposed algorithm is compared with five state-of-the-art many-objective evolutionary algorithms on a large number of test problems with various characteristics. Experimental studies demonstrate that the proposed algorithm has competitive performance on many-objective optimization problems.