云模型改进惯性权重的混沌交替粒子群算法  

Chaos Alternation Particle Swarm Optimization Algorithm Improved Inertia Weight Based on Cloud Model

作　　者：李纪真[1] 孟相如[1] 崔文岩[1] 杨婷 

机构地区：[1]空军工程大学信息与导航学院 [2]解放军95133部队

出　　处：《火力与指挥控制》2016年第5期56-61,共6页Fire Control & Command Control

基　　金：国家自然科学基金资助项目(61201209;61401499)

摘　　要：标准粒子群算法通过线性减小惯性权重系数来调整寻优性能,但缺乏智能化机制易导致算法后期产生早熟或陷入局部最优而产生僵局。针对这一问题,提出一种基于云模型改进惯性权重的混沌交替粒子群优化算法。根据粒子迭代变化关系,采用云模型理论对惯性权重ω进行智能化调整,以平衡其全局和局部搜索能力,防止算法产生局部僵局;另外,判定粒子稳定性,对于可能陷入局部僵局的稳定粒子进行混沌扰动,促使其跳出僵局进而向最优位置更新。实验与分析表明,基于云模型改进惯性权重的混沌交替粒子群优化算法能够跳出局部僵局且具有较高的寻优精度,算法接近完全收敛时的平均迭代次数,较现有相关研究分别降低了13.73%~20.11%。The optimization performance of the standard particle swarm optimization algorithm isadjusted by reducing the inertia weightlinear,which lack of intelligent mechanism and easy to bringinto the prematurity and local stalemate in the evening of the algorithm. A chaos alternation particleswarm optimization algorithm improved the inertia weight based on cloud model is proposed to solvethese problems. The inertia weight 棕of the particle swarm optimization algorithm is adjusted by cloudmodel intelligently according to the iterative transformation of the particles,and the whole and localsearching capabilities of the particle swarm optimization algorithm get balanced,and prevent it into thelocal stalemate. In addition,determine the stability of the particles,and do chaos disturbance to thestable particles which maybe bring into the local stalemate,make it jump out form the local stalemateand close to the optimization position further. It shows from the experiment and analysis that the chaosalternation particle swarm optimization algorithm improved the inertia weight based on cloud model canbe able to jump out from the local stalemate with a higher optimization precision,and the averageiterative numbers are reduced by 13.73 豫~20.11 豫than other researches when the algorithm getsabsolutely convergence.

关 键 词：粒子群 云模型 惯性权重 稳定粒子 混沌交替 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]