A new many-objective evolutionary algorithm based on self-adaptive differential evolution

To improve the performance of the existing multi-objective evolutionary algorithms (MOEAs), we propose a new self-adaptive differential evolution algorithm for solving many-objective optimization problems (MOPs). To address the challenges in many-objective optimization, new selection strategy and density estimation method are designed to improve the performance of the elite MOEA model used by several exiting MOEAs. In addition, new mutation strategy and parameter adaptive method of DE are proposed to enhance the convergence ability of the evolution strategy utilized in MOEAs. Experimental results on ZDT and DTLZ test problems show that, the proposed algorithm, named SDEMO, is able to find much better spread of solutions with better approximating the true Pareto-optimal front compared to six state-of-the-art MOEAs.