基于惯性权重自适应改变的混沌粒子群算法  

Chaos Particle Swarm Algorithm Based on the Adaptive Changing Inertia Weight

作　　者：郝继升[1] 朱文婷[1] 袁倩[1] 雷珍 

机构地区：[1]延安大学数学与计算机科学学院,陕西延安716000

出　　处：《河南科学》2015年第6期942-945,共4页Henan Science

基　　金：陕西省教育厅项目(2013JK1124);陕西省高水平大学建设专项资金资助项目(2012SXTS06)

摘　　要：粒子群算法是美国学者受鸟类觅食行为启发提出的一种群体优化算法,在迭代后期易早熟收敛.为此利用混沌运动的随机性、规律性和遍历性的特点对粒子群算法进行优化,提出了一种惯性权重自适应改变的混沌粒子群算法.在算法中对惯性权重进行调整加快算法前期收敛速度,而且加入了变异操作以帮助粒子后期跳出局部极小.最后用测试函数进行仿真实验,结果表明该算法收敛快,寻优能力强,寻优精度高.Particle swarm algorithm proposed by American scholars who inspired by the foraging behavior of birds is a colony optimization algorithm,which tends to premature convergence in later iterations. By using randomness, regularity,ergodicity of chaotic movement to optimize particle swarm algorithm,a chaos particle swarm algorithm with adaptive changing inertia weight is proposed. The inertia weight is adjust ed to accelerate the convergence of former iterations. Furthermore,joining mutation in the algorithm helps avoiding trapping to local minima. Simulation results of the test functions show that the algorithm has fast convergence and strong optimization capability and higher optimization precision.

关 键 词：粒子群 混沌 自适应 变异 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]