【摘要】 针对高维优化问题提出基于改进的微粒群及模拟退火方法的SA-PSO算法。算法引入了随机选择、小邻域赋值、极值扰动,以及渐进搜索等策略,能有效解决微粒规模与优化问题维数之间的矛盾,加快收敛速度并防止陷入局部最优解。经实例应用检验,SA-PSO算法能够达到预期的效果,开辟了求解高维优化问题的新途径。更多还原

【Abstract】 A combined SA-PSO algorithm is proposed and developed based on traditional Particle Swarm Optimization(PSO) and Simulated Annealing(SA) for dealing with the supper dimensions problem.Some strategies and techniques have been adopted in the combined algorithm such as stochastic selection,evaluate constraint,disturbance model and successive approximate searching.The character of this algorithm is calculating the complex overrun dimensions problem with a small particle scale and enhancing the optimizing efficiency by accelerating convergence speed and jumping out the local optimal result.The research indicated that the proposed algorithm is efficient and it is possible for dealing with the multi-dimension problems.更多还原