摘要
针对传统算法在求解物流配送中心选址问题时容易陷入局部最优解和寻优效果不够理想的缺陷,提出了一种改进的粒子群算法。该算法通过引入领域均值来反映粒子间合作与竞争的隐性知识,使粒子种群的多样性和算法的全局搜索能力得到改善;利用边界缓冲墙对超越边界的粒子进行缓冲,使算法的收敛速度和寻优精度有明显的提高。仿真实验结果表明,该算法比传统方法具有更好的性能,特别是当物流需求点的数量很大时,该算法的优越性更加明显。
Aiming at the problems of traditional algorithms for solving the logistics distribution center location problem with falling into local optimal solution and the bad optimization effects,this paper presented an improved particle swarm optimization algorithm.By utilizing the field means to reflect the tacit knowledge of cooperation and competition during the process of optimization,it improved the diversity of particle population and the global search capability of algorithm.By using the boundary buffer wall,it buffered the particles beyond the boundary,and significantly improved convergence speed and optimization accuracy.Simulation results show that the improved algorithm has better performance than traditional methods,especially when the number of logistics demand point are very large,the improved algorithm has more obvious superiority.
出处
《计算机应用研究》
CSCD
北大核心
2012年第12期4489-4491,共3页
Application Research of Computers
基金
国家自然科学基金资助项目(71171135)
上海市(第三期)重点学科资助项目(S30504)
四川省系统科学与企业发展研究中心资助项目(XQ12C09)
绵阳师范学院青年基金资助项目(2012A06)
关键词
粒子群算法
配送中心
选址模型
物流
智能优化
particle swarm optimization（PSO）
distribution center
location selection
logistics
intelligence optimization
