多策略融合的改进万有引力搜索算法  

Improved gravitational search algorithm with multi-strategy fusion

作　　者：樊康生 杨光永[1] 吴大飞 汪军 徐天奇[1] Fan Kangsheng;Yang Guangyong;Wu Dafei;Wang Jun;Xu Tianqi(School of Electrical&Information Technology,Yunnan Minzu University,Kunming 650500,China)

机构地区：[1]云南民族大学电气信息工程学院,昆明650500

出　　处：《计算机应用研究》2023年第12期3592-3598,共7页Application Research of Computers

基　　金：国家自然科学基金资助项目(61761049,61261022);2023年度云南省教育厅科学研究基金资助项目(2023Y0502);云南民族大学2022年硕士研究生科研创新基金资助项目(2022SKY006)。

摘　　要：为解决传统万有引力搜索算法(GSA)易陷入局部最优和开发能力弱等问题,提出了一种多策略融合的改进万有引力搜索算法(MFGSA)。首先,提出动态调整引力常数G的更新策略,以增强算法的探索能力和收敛精度;其次,为保留粒子的多样性,提出了基于对称思想的粒子越界处理策略,以提高算法的收敛精度;为适应前两个策略,还引入精英思想,用最优粒子改善最差粒子位置策略,以避免算法陷入局部最优;同时,提出了自适应因子更新粒子速度和位置策略,以提高算法的收敛速度。为验证改进算法的性能,将改进算法与传统万有引力搜索算法和其他四种改进万有引力搜索算法在10个基准函数上进行了对比实验,结果表明MFGSA在收敛速度、搜索精度方面优势较大,表明MFGSA性能的优越性。To solve the problems being easy to fall into local optimum and weak development capability of traditional universal gravitational search algorithm(GSA),this paper proposed an improved gravity search algorithm with a multi-strategy fusion(MFGSA).Firstly,in order to enhance the exploration capability and convergence precision of the algorithm,the improved algorithm used an update strategy for dynamically adjusting the gravitational constant G.Secondly,in order to preserve the diversity of particles and improve the convergence precision,the improved algorithm proposed a particle crossing processing strategy based on symmetry idea.To accommodate the first two strategies,the improved algorithm used the elitist strategy which was introduced to improve the position of the worst particles with the optimal particles to avoid the algorithm falling into local optimization.At the same time,the improved algorithm proposed a self-adaptive factor to update particle velocity and position strategy to improve the convergence speed of the algorithm.This paper designed a few compared experiments with the traditional universal gravitation search algorithm and other four improved universal gravitation search algorithms on 10 benchmark functions to verify the performance of the improved algorithm.The results show that MFGSA has great advantages in convergence speed and search accuracy,which proves the superiority of MFGSA performance.

关 键 词：多策略融合 改进万有引力搜索算法 引力常数 自适应因子 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]