Research And Improvement Of Swarm Intelligence Optimization Based On Dynamic Multi-swarm Strategy

Swarm intelligent optimization algorithm has been extensively used in various fields such as economy and engineering due to its advantages of simple implementation,fast convergence and high robustness.However,early swarm intelligence optimization algorithms often have problems such as premature convergence,lack of population diversity and poor convergence accuracy,etc.,especially when solving multi-peak and complex functions,the algorithm is often difficult to give consideration to population diversity and convergence accuracy.Therefore,on the basis of previous studies,this paper combines the dynamic multipopulation strategy with Particle Swarm Optimization(PSO)and Pigeon-inspired Optimization(PIO)to solve the premature convergence problem of Swarm intelligence optimization algorithm,and proposes three improved algorithms.The main work are as follows:1.This paper propose a dynamic multi-swarm pigeon-inspired optimization.In DMS-PIO,all individuals are divided into several small sub-swarms at the early stage of evolution,and the diversity of the population is maintained through the independent evolution of several subswarms.In order to realize the sharing of dominant information among populations,the subswarms were reorganized periodically.Experimental results show that the improved algorithm can effectively improve the problem that PIO algorithm is prone to premature convergence.2.A dynamic multi-swarm global particle swarm optimization is proposed.In DMS-GPSO,the whole evolutionary process is divided into two stages.In the early days of the search,the entire population is divided into two sub-swarms,one called global sub-swarms and the other called dynamic sub-swarms,which focus on local and global search respectively.The storereset operation is used to improve the search capability of the algorithm.Experiments show that the dynamic multi-swarm strategy and storage-reset strategy effectively improve the ability of the algorithm to solve multimodal and complex functions.3.A dynamic multi-swarm particle swarm optimization based on elite learning is proposed.The algorithm divides the whole population into several dynamic sub-swarms and one following sub-swarm according to the fitness value of the particle.On this basis,the random grouping method based on elite learning strategy is used to improve the global and local search ability of the algorithm.Experiments show that the convergence accuracy of the algorithm is improved by multi-swarm strategy on the premise of ensuring population diversity.