Font Size: a A A

Research And Application Of Adaptive Chaos Particle Swarm Optimization Based On Predatory Search Strategy

Posted on:2014-06-17Degree:MasterType:Thesis
Country:ChinaCandidate:T WangFull Text:PDF
GTID:2268330392463695Subject:Applied Mathematics
Abstract/Summary:PDF Full Text Request
This paper is devoted to the study of the hybrid particle swarm optimization.The object matter and innovation as follows:In this paper, through the introduction of predatory search strategy, chaotic mutation strategyand adaptive inertia weight strategy, a new adaptive chaos particle swarm optimization based onpredatory search strategy is proposed. The algorithm uses fitness value to regulate globaloptimizing and local optimizing by introducing the predatory search strategy. It strengthensoptimizing near a better solution and improves the accuracy of the algorithm. At the same time,chaos mutation search strategy is introduced: On the one hand, a chaotic sequences is used toinitialize the particle position and velocity and a certain percentage of the outstanding initialindividuals are selected. On the other hand, When the particles search times don’t find bettersolutions, chaos optimizes the best position of the swarm, gets the best solution in the sequenceby the chaotic searching, calculating the fitness value of the sequence and comparing it with thefitness value of the current best solution. If it is better than the current best solution, the bestchaotic sequence replace the current optimal particle, or replace any particle with the best chaoticsequence according to mutation probability. So it will increase the diversity of population, helpsthe particle to jump out of local optimum and enhances global exploration abilities of particles.In order to improve the convergence speed of particle and balance the global search and localsearch, the dynamic adaptive selection strategy of inertia weight. is used. Finally, simulating andanalyzing the optimization performance of this algorithm by5test functions. It shows that thealgorithm has the global convergence ability and the local search ability, and it improvesconvergence speeds of particles and the accuracy of optimal value.
Keywords/Search Tags:Particle Swarm Optimization, predatory search, chaos, cube map, adaptive, mutation, jump out of local optimal
PDF Full Text Request
Related items