Research And Improvement Of Multi-objective Particle Swarm Optimization Algorithm

As a typical swarm intelligence optimization algorithm,the particle swarm optimization algorithm has many advantages,such as simple model,convenient operation,easy realization and good robustness,it also has profound intelligence features,all of these advantages provide a new idea for solving the complex problems and shows vigorous vitality,also attract a large number of experts and scholars to explore it.But the particle swarm optimization itself has some disadvantages such as slow convergence speed and easy premature convergence.Therefore,the improvement of particle swarm optimization is still an important issue.In this paper,an improved multi-population particle swarm optimization algorithm and an improved multi-objective particle swarm optimization(PSO)algorithm are proposed for the different number of targets to be optimized.The main results of this paper are as follows:(1)The research status of the particle swarm optimization algorithm is analyzed and the advantages and disadvantages of the particle swarm optimization algorithm are given,and the algorithm principle and mathematical model of the particle swarm optimization algorithm are analyzed too.(2)In order to improve the quality of the initial solution,improve the convergence speed of particle swarm optimization and enhance the ability of PSO to jump out of the local optimal solution,an improved multi-population particle swarm optimization algorithm was proposed in this paper.This algorithm decomposes the population into three subgroups,and then analyzes the population initialization,analyzes the judgment of premature convergence and analyzes the mining and detecting ability of the population and the cooperation and learning among the subgroups,and gives the corresponding improvement application.At the same time,for the premature convergence problem,a new particle update strategy was proposed in this paper.Can effectively jump out of local optimal.(3)The improved multi-population particle swarm optimization algorithm proposed in this paper is applied to the lobe control of the pattern synthesis of the array antenna pattern.The test results show that the improved algorithm can effectively solve the problem.(4)In this paper,an improved multi-objective particle swarm optimization algorithm is proposed to solve the problems of low precision,slow convergence and local optimality in the multi-objective optimization.Firstly,the algorithm uses uniform initialization strategy to initialize the whole population,and improves the uniformity of the original solution distribution.Secondly,unlike most current algorithms to maintain external archive,this algorithm reuses and enhances the particles which beyond the external archive size and improves the diversity of the solution distribution.Finally,in order to solve the problem that particle swarm optimization algorithm is easy to fall into local optimization,the improved algorithm establishes the particle information archives to store all the information of the particles,and compare the update of several generations of particles in these particle information archives to judge the state of the particle and take different update strategies,so as to make the particle update more efficient and improve the overall performance of the algorithm to solve complex multi-objective problems.By comparing the improved multi-objective particle swarm optimization algorithm with other classical algorithms,the improved algorithm proposed in this paper has better overall performance advantage in solving multi-objective problems.