Improved Multi-Objective Particle Swarm Optimization Algorithm With Adaptive Strategies

Abstract

For solving multi-objective optimization problems, we propose a multi-objective particle swarm optimization algorithm based on Adaptive Strategies (ASMOPSO). The algorithm evaluates the diversity of the external population in each iteration, and adaptively chooses whether to perform mutation operations on the external population and choose different particle population update methods according to the evaluation value. The method cooperates with crowding distance and non-dominant sorting. In each round of iteration, different particles before and after sorting are changed to different scales to avoid the algorithm from falling into premature convergence and falling into the local optimum. It can be found through experimental simulation that the performance of the algorithm is better. Compared with evolutionary algorithm and general multi-objective particle swarm algorithm, it is greatly improved, and it shows better results on multiple test functions.