Solving Many-Objective Optimization Problems via Multistage Evolutionary Search

With the increase in number of optimization objectives, balancing convergence and diversity evolutionary multiobjective becomes more intractable. So far, a variety algorithms have been proposed to solve many-objective problems (MaOPs) with than three objectives. Most existing algorithms, however, find difficulties simultaneously counterpoising during whole process. To address issue, this paper proposes MaOPs via multistage search. be specific, two-stage algorithm is developed, where are highlighted different search stages avoid interferences between them. The first stage pushes multiple subpopulations weight vectors converge areas Pareto front. After that, nondominated solutions coming from each subpopulation selected for generating new population second stage. Moreover, environmental selection strategy designed balance close This evenly divides objective dimension into intervals, then one solution having best interval will retained. assess performance algorithm, 48 benchmark functions 7, 10, 15 objectives used make comparisons five representative algorithms.