摘要
为克服粒子群优化算法(Particle Swarm Optimization,PSO)存在的缺陷,提出基于Logistic映射的自适应变尺度混沌粒子群优化算法(Adaptive Chaos PSO,ACPSO)。采用混沌方法对粒子进行初始化;根据不同状态下粒子适应值的大小对惯性权重采取不同的调整方法;异步变化的学习因子使粒子随着迭代步数的增加,避免粒子发生早熟收敛现象;当粒子陷入局部最优时,对部分较优粒子采用变尺度混沌局部优化策略。为了检验算法的有效性,将该算法与3种有代表性的算法进行比较,结果表明该算法收敛速度快,求解精度高。
To overcome the shortcomings of Particle Swarm Optimization （PSO）, an Adaptive Mutative Scale Chaos Particles Swarm Optimization （ACPSO） based on Logistic Mapping was proposed. The chaos method was used to initialize the particles. The adjustment method of the inertia weight depended on the particle＇s fitness; it could avoid premature convergence for the particles. When the particles fell into the local optimum, mutative scale chaos optimization strategy was adopted to adjust the optimal particles. To test the effectiveness of the algorithm, three representative algorithms were compared with. The results show that the algorithm has high convergence speed and high precision.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2017年第10期2241-2246,共6页
Journal of System Simulation
基金
国家自然科学基金(61503124)
高等学校重点科研项目(15A520018)
关键词
粒子群优化
变尺度
混沌优化
自适应
学习因子
惯性权重
PSO
mutative scale
chaos optimization
adaptive adjustment
learning factor
inertia weight
