摘要
针对综合学习粒子群算法后期收敛速度慢、一旦所有粒子陷入局部最优,则无法跳出等缺陷,提出免疫综合学习粒子群优化(ICLPSO)算法。ICLPSO算法引入人工免疫系统中的克隆选择机制,利用克隆复制、高频变异、克隆选择等操作,增加种群的多样性,提高算法的收敛速度,利用柯西分布较宽的两翼分布特性进行精英粒子学习以进一步增强粒子逃离局部极值及多峰函数优化问题全局寻优能力。针对标准测试函数的仿真结果表明,与其他改进粒子群算法相比,ICLPSO算法收敛速度快,求解精度更高。
Convergence of the comprehensive learning particle swarm optimization(CLPSO)algorithm is relatively slow at the late stage of evolution.Once all particles trapped in local optimum,the algorithm can not jump out of the local optimum.This paper proposed immune comprehensive learning particle swarm optimization(ICLPSO)algorithms.The algorithm introduced clonal se-lection mechanism in artificial immune system.Using of clonal copy,hypermutation and clonal selection,it increased the diversi-ty of the population,improved the convergence rate and enhanced the ability of escape from the local optimum and multi-mode op-timization ability of global optimization.Using the elitist learning strategy,the ability to escape from local optimia is further en-hanced.Experiments on several benchmark functions verify the effective of the proposed algorithm.
出处
《计算机应用研究》
CSCD
北大核心
2014年第11期3229-3233,共5页
Application Research of Computers
基金
国家自然科学基金资助项目(61174140)
国家教育部博士点基金资助项目(20110161110035)
湖南省自然科学基金资助项目(11jj4049)
关键词
综合学习粒子群算法(CLPSO)
人工免疫系统
精英学习
函数优化
comprehensive learning particle swarm optimization algorithm
artificial immune system
elitist learning
function optimization