An Improved NSGA-II algorithm for multi-objective nonlinear optimization

Non-dominated sorting genetic algorithm (NSGA-II) has been widely used in the field of multi-objective optimization after introducing sorting algorithm, crowding degree operator and elite strategy, but there are also problems such as uneven distribution of individual and Pareto efficiency. The algorithm is improved for these defects. Firstly, a cumulative non-dominated ranking assignment strategy is proposed to improve the crossover operator. The DE algorithm with adaptive parameters is used to improve the initial population and the population diversity; Then introduce the idea of nonlinear algorithm to improve the local search ability and further improve the quality of Pareto optimal solution. Finally, six benchmark multi-objective functions are used for testing. The experimental results show that the improved NDE-NSGA-II algorithm has better distribution, stability and higher search solution.