Research And Applications Of Particle Swarm Optimization

Particle swarm optimization (PSO) is a population-based, self-adaptive search optimization technique, and was originally introduced by J.Kennedy and R.C.Eberhart in 1995. As a kind of swarm intelligence algorithm, PSO is simple in principles, few in parameters, fast in convergence rate and so on. It has proven to be a powerful global optimization method. So it has attached great importance to researchers home and abroad and found applications in many areas since it was introduced. But neither the theoretical analysis nor applications are not completely mature, yet have many questions to be researched.This paper does some research on the improvements and applications of PSO. The detailed jobs are as follows:First, this paper attempts to study the application of PSO in variational problems. In this paper, an optimization model is found by analyzing the variational problems, and the fitness function for PSO is established in the model. So the variational problems are solved by PSO successfully. And the application field of PSO is widened. Then, according to the variational principle, PSO is used to solve the Second-order linear ordinary differential equations problem in the same way.Second, this paper focuses on solving the transportation problem by PSO. It proposes a method to product initial feasible solutions of transportation problem, on the basis of its special constrained conditions, and the unfeasible solutions need not to care any more. At the same time, a method namely GAPSO algorithm to solve transportation problem is presented by combining the genetic algorithm with particle swarm optimization. Finally, examples are illustrated the proposed approach to verify its effectiveness and practicality.Third, this paper does research on the ability of local search and converging speed of PSO in later period. Based on the hierarchical genetic algorithm, a new hierarchical particle swarm optimization was proposed. Experimental results on several benchmark functions indicate that the hierarchical particle swarm optimization increases the speed of convergence and enhances the ability of local search.In the end, after summarizing the whole paper, we pointed out some problems to be solved in the future.