An Improved Selection Method Based on Crowded Comparison for Multi-Objective Optimization Problems in Intelligent Computing

The main method of dealing with multi-objective optimization problems (MOPs) is the improvements of non-dominated sorting genetic algorithm II (NSGA-II), which have obtained a great success for solving MOPs. It mainly uses a crowded comparison method (CCM) to select the suitable individuals for enter the next generation. However, the CCM requires to need calculate the crowding distance of each individual, which needs to sort the population according to each objective function and it exhausts a lot of computational burdens. To better deal with this problem, we proposes an improved crowded comparison method (ICCM), which combines CCM with the random selection method (RSM) based on the number of selected individuals. The RSM is an operator that randomly selects the suitable individuals for the next generation according to the number of needed individuals, which can reduce the computational burdens significantly. The performance of ICCM is tested on two different benchmark sets (the ZDT test set and the UF test set). The results show that ICCM can reduce the computational burdens by controlling two different selection methods (i.e., CCM and RSM).