Convergence Analysis And Improvement Reseach Of Standard Particle Swawrm Optimization Algorithm

Traditional optimization algorithm could not figure out some complicated optimization problems owing to the fact that more and more optimization problems appeared in real life.Therefore,some novel heuristic optimization algorithms are proposed,such as genetic algorithm,simulated annealing algorithm,gravitational search algorithm,ant colony algorithm and so on.These algorithms are basically to simulate or reveal the natural phenomena or processes and development,their ideas and content related to mathematics,physics,biological evolution,artificial intelligence,neuroscience and statistical mechanics and so on.The Particle swarm optimization(PSO)is one of them,with simple regulations,few adjustable parameters,easily achieved,quick convergence rate,and a lot measures can avert PSO trapped in the local optimum,also widely used in project applications,with its excellent properties,the present paper make a further study to PSO,the main work was as follows:1.Based on the probability theory,the transition probabilities of the particles in the standard particle swarm optimization algorithm are calculated,and analyzed PSO convergence on the basis of transition probability,then proved that the standard PSO algorithm is convergence with probability 1 under certain condition.2.A new improved particle swarm optimization(IPSO)algorithm is proposed to ensure that IPSO algorithm is convergence with probability 1.In order to balance the exploration and exploitation abilities of IPSO algorithm,we propose the exploration and exploitation operators in IPSO algorithm.Finally,IPSO algorithm is tested on 13 benchmark test functions and compared with the other algorithms published in the recent literature.The numerical results confirm that IPSO algorithm has a better performance in solving nonlinear functions.