摘要
针对PSO易收敛于局部最优的缺点,提出了运用免疫小生境思想来改进PSO。该算法在初始化时,运用正交的思想,使得粒子分布均匀;且在进化时,通过对每个粒子做免疫变换,使得每个粒子扩展成为在一个区域寻找最优值,提高了粒子的多样性,避免了局部最优;并且在变换时,每隔几代才进行免疫变换。这样在保证粒子多样性的基础上减少了运算量,提高了收敛速度。并在MATLAB环境下对Ackley函数、Schaffer函数、Griewank函数、Rastrigrin函数四个多峰函数进行了仿真验证,实验结果表明,改进的PSO算法能够有效地达到全局最优。
Due to the premature convergence of PSO,immunity niche concept is adopted to improve the performance of PSO.Orthogonal principle is applied to the initialization of particle,so particles are evenly distributed.During the evolution,every particle is extended to a region by immunity operation,so the algorithm can improve the diversity and has the capability of avoiding premature convergence.Since the immunity operation is only performed every several generations,The convergence velocity is improved.And it is verified by Ackley function,Shcaffer function and Grievank function in MATLAB,the result shows that the improved PSO can realize global optimization effectively.
出处
《计算机工程与应用》
CSCD
北大核心
2008年第8期95-97,共3页
Computer Engineering and Applications
