Font Size: a A A

The Research And Improvement Of Particle Swarm Optimization

Posted on:2007-09-29Degree:MasterType:Thesis
Country:ChinaCandidate:J Y LiuFull Text:PDF
GTID:2178360185485954Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Since the 1980s, intelligent optimization algorithm such as neural network, GA has been developed through the simulation of nature and social process and it presents a new approach for optimization methods. Particle swarm optimization (PSO) is inspired by social behavior of bird flocking or fish schooling. It is a population-based, self-adaptive search optimization technique. PSO is simple in concept, few in parameters, and easy in implementation. As a kind of intelligent algorithm, it can be used to solve various optimization problems and shows great potential in practice. Now, it has been widely applied in many other areas, such as function optimization, artificial neural network and fuzzy system control.In this paper, the basic principles of PSO are introduced. The research progress on PSO algorithm is summarized such as parameter selection and design, population topology, hybrid PSO algorithm etc. Linearly decreasing inertia weight (LDW) and constriction factor model (CFM) are two standard PSO algorithms. Both of them efficiently improve the performance of basic PSO, but they still have the weakness of premature convergence when they are used in multi-dimension problems.Further, an improved algorithm is proposed using the characteristics of the flight of geese for reference. The improved algorithm has superiority over PSO; for one thing, it keeps the population various by ordering all particles and making each particle fly following its anterior particle; for another thing, it strengthens cooperation and competition between particles by making each particle share more useful information of the other particles. Several benchmark functions are tested and the experimental results show that the new algorithms not only significantly speed up the convergence, but also effectively solve the premature convergence problem.
Keywords/Search Tags:particle swarm optimization, swarm intelligence, inertia weight, constriction factor, flight of geese
PDF Full Text Request
Related items