Optimization Of Multi-objective Particle Swarm Optimization Based On Game Mechanism And Mixed Strategy

In real life,many optimization problems have multiple goals and must be optimized at the same time.Traditional techniques can no longer solve the problem of multi-objective optimization.The meta-heuristic algorithm provides new ideas for solving multi-objective problems,and particle swarm optimization is the representative algorithm of meta-heuristic algorithm,the particle swarm algorithm idea is derived from the behavior of bird predation in nature.Although there are many multi-objective particle swarm optimization algorithms,due to particle swarm's special leader mechanism and premature problems,the multi-objective particle swarm optimization algorithm has problems of convergence and diversity deviation in the process of optimization.In view of the above problems,this paper proposes an improved method based on in-depth study of multi-objective particle swarm optimization algorithm,the main work is as follows:(1)In order to improve the problem of multi-objective particle swarm optimization,the lack of diversity of optimization solutions and the convergence problem of the algorithm,a multi-objective particle swarm algorithm with game mechanism is proposed.The algorithm does not require an external set to store the global optimal value and individual optimal value.Firstly,the characteristics of chaos mapping are used to improve the convergence speed of the algorithm to avoid the algorithm falling into a local optimum.Secondly,the crowded distance and game mechanism maintain the diversity of the algorithm.Experimental results show that compared with six algorithms on three series of test functions,the proposed algorithm has good performance in maintaining population diversity and convergence speed.(2)In order to balance the convergence and diversity in the search process,a new hybrid strategy multi-objective particle swarm optimization algorithm is proposed.This method proposes a simplified leader-oriented particle swarm optimization algorithm to accelerate the convergence speed of particles;Second,make full use of the learning mechanism of the particle swarm algorithm to extract particle velocity information,a new elite-defined measurement method and an analysis method of decision variables based on enhanced diversity,which enhances the diversity of the solution obtained by the algorithm;adaptive dual leadership Choose a strategy to avoid the algorithm falling into a local optimum.The experimental results on benchmark examples show that the performance of the hybrid multi-objective particle swarm optimization algorithm is better than other multi-objective particle swarm optimization algorithms.Compared with evolutionary algorithms based on decomposition and control,the ability to maintain diversity in high-dimensional target space is improved.