摘要
微粒群算法(PSO)是一种随机群体优化算法,相对于遗传算法等其它的进化算法,它模型简单、操作参数少、智能程度高、运算速度快,已受到许多相关领域学者的关注与研究。但是,标准微粒群算法在寻优过程中往往陷入局部最优解,而不是全局最优解。在研究均匀设计与惰性变异的基础上,提出了改进的微粒群算法(UMPSO)。该算法利用均匀设计的思想来确定算法的初始粒子,以使其均匀分布于解空间,从而使算法以更高的概率、更快的速度找到全局最优解;在进化过程中,对惰性粒子以概率为1进行随机变异,则能够更好地保证微粒群的多样性。仿真结果表明,与标准的PSO相比,UMPSO的寻优精度更高、寻优速度更快。
Particle swarm optimization （PSO） is a population-based stochastic optimization technique. It shares many similarities with evolutionary computation techniques such as Genetic Algorithms （GA）. But compared with GA, it has simpler model, fewer parameters, higher intelligence, faster computation, which make it attractive to some researchers. Otherwise, the standard Particle Swarm Optimization often doesn＇t find global optimal solution, but the local solution. Based on research on uniform design and inertia mutation （UMPSO）, an advanced particle swarm optimization was proposed, which initializes population with uniform design to make them distribute in the problem place evenly. This makes UMPSO find the global optimal solution with higher probability and faster velocity. During the process of evolution, UMPSO carries out mutation for the inertia particles with the probability of I, which keeps population diverse. Experimental results show that UMPSO can find optimal solution more precisely and faster than the standard PSO does.
出处
《系统仿真学报》
EI
CAS
CSCD
北大核心
2008年第10期2584-2588,共5页
Journal of System Simulation
基金
国家自然科学基金重点项目(60534040)
广东省自然科学基金自由申请项目(05001819)
关键词
微粒群算法
均匀设计
惰性变异
进化计算
遗传算法
particle swarm optimization
uniform design
inertia mutation
evolutionary computation
genetic algorithm
