Improved NSGA-II with Hierarchical Retention and Interval Population Expansion

In order to overcome the shortcomings of the classical fast non-dominated sorting genetic algorithm with elitist strategy (NSGA-II), such as the uneven pareto front distribution and poor distribution in local congested areas, an improved NSGA-II algorithm based on adaptive hierarchical retention and interval population expansion strategies is proposed. At the early stage of the population evolution, an adaptive hierarchical retention strategy is applied to replace the exclusion mechanism to get an expanded range of individuals selection and improve the diversity of the population. At the last stage of the population evolution, an interval population expansion strategy is provided to expand the contemporary optimal frontier individuals to reduce the sparsity of the swarm distribution aiming to improve the comprehensive performance of the population. The experiment results based on six selected benchmark functions show that the proposed algorithm gets a better convergence and is superior to the compared algorithms in the terms of comprehensive and distribution values.