摘要
针对万有引力搜索算法对一些优化问题的搜索精度不高,容易出现早熟的问题,提出了改进的万有引力搜索算法.该算法借鉴生物界中雁群的飞行特征和加权平均法,扩大了搜索范围,加强了粒子间的合作与竞争.通过对6个基准函数的仿真测试,MATLAB仿真结果表明新算法能更有效地提高全局搜索能力.
Aiming at the problems that gravitational search algorithm easily falls into premature convergence and has bad performance in search accuracy,improved gravitational search algorithm（IGSA）is put forward.This new method,combining the flight characteristics of wild geese flock and the method of weighted mean,expands the search scope and strengthens the cooperation and competition between the particles.The simulation test of the 6benchmark functions by MATLAB confirms that the improved algorithm efficiently improves the global search ability.
出处
《沈阳大学学报（自然科学版）》
CAS
2014年第6期468-472,共5页
Journal of Shenyang University：Natural Science
基金
国家自然科学基金资助项目(61074099)
关键词
万有引力搜索算法
雁群飞行特征
加权平均法
数值函数优化
gravitational search algorithm
the flight characteristics of wild geese flock
the method of weighted mean
numerical function optimization
