摘要
针对万有引力搜索算法(GSA)在处理一些函数优化问题时容易出现早熟和搜索精度不高的缺点,提出了一种改进万有引力搜索算法。该算法结合小生境技术中的共享机制,通过反映粒子之间相似程度的共享函数来调节群体中各个粒子的适应度,提高了万有引力搜索算法中粒子的多样性。4个常用测试函数的仿真实验结果表明:与万有引力搜索算法相比,改进万有引力搜索算法在求解函数优化问题时具有更好的优化性能。
Aiming at the problems that gravitational search algorithm(GSA) easily falls into premature convergence and has bad performance in search accuracy,an improved gravitational search algorithm is put forward.With a combination of the sharing mechanism of niche technology,the algorithm adjusts the particle fitness to improve the particle diversity using the share function which reflects the similarity degree among particles.The simulation results of four nonlinear benchmark functions showthat the improved gravitational search algorithm has much better optimization performance in solving various nonlinear functions than basic gravitational search algorithm.
作者
杨元荣
YANG Yuanrong(College of Science, Hohai University, Nanjing 211100, Chin)
出处
《系统仿真技术》
2018年第1期78-82,共5页
System Simulation Technology