A Many-Objective Evolutionary Algorithm Based on Non-Linear Dominance

With the increase in number of objectives, non-dominated solutions will also sharply. The sorting method based on traditional Pareto dominance is not sufficiently distinguishable from and cannot provide enough selection pressure when population size small. In this article, a new non-linear (NLD) proposed. main motivation perspective storage solutions. small difference between each component as large possible, so part first quadrant, second, fourth quadrant near becomes dominant interval, except for distance too far defined which construct parabolic shape non-dominant interval. Based relationship, authors propose dominated many-objective evolutionary algorithm (NLDEA), can solve irregular front. Experiments show that NLDEA competitive with most advanced methods various scalable benchmark problems.