摘要
基于对不同粒子群算法(PSO)中惯性权重、全局收敛性、收敛精度和速度的分析,提出了一种新的全局最优值自适应变化的粒子群算法(LAPSO)。并采用该方法对三种不同的基准函数进行了测试,将LAPSO测试结果与典型的收敛粒子群算法(LKPSO)和扩散粒子群算法(LWPSO)进行了比较。结果表明:自适应粒子群算法具有收敛速度快、进化精度高的特点,是一种新型全局收敛粒子群算法。
Based on the analysis of inertia weight, global convergence and convergent speed as well as accuracy, a novel adaptive particle swarm optimization (LAPSO) was proposed. This landscape method was applied to investigate three different Benchmark functions. Compared with the experimental results of LWPSO and LKPSO, the results show that LAPSO is more performance in speed and accuracy convergence. Therefore, LAPSO is a global particle swarm optimization.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2009年第9期2582-2585,共4页
Journal of System Simulation
基金
教育部重点科学研究项目(105087)
国防应用基础研究基金项目(A1420061266)
关键词
粒子群算法
惯性权重
全局最优值
局部收敛
particle swarm optimization
global optimality
convergence
local convergence