基于聚类-粒子群算法的冷链物流配送中心选址分析  

Site Selection Analysis of Cold Chain Logistics Distribution Center Based on Clustering Algorithm and Improved Particle Swarm Optimization

作　　者：徐超毅[1] 刘涛 Xu Chaoyi;Liu Tao(School of Economics and Management,Anhui University of Science&Technology,Huainan City,Anhui Province 232001)

机构地区：[1]安徽理工大学经济与管理学院,安徽淮南232001

出　　处：《黄河科技学院学报》2023年第8期19-27,共9页Journal of Huanghe S&T College

基　　金：郑州市第六届中青年骨干教师项目(zzjs201902)。

摘　　要：针对冷链市场需求不断扩大,冷链物流产业却无法快速升级的情况,终端配送企业为了提高企业效益,需要在保证满足客户需求的同时,构建以成本最小化为目标的冷链配送中心选址模型。为解决上述问题,首先使用Mean-shift聚类算法对未知形状、未知特征的需求点进行处理,确保配送中心选址在空间分布上的合理性,同时降低运算难度;其次引入反向搜索、变异和交叉策略对传统粒子群算法(PSO)进行改进,使用改进粒子群算法(IPSO)对模型进行求解。相关实例分析结果表明,使用IPSO算法的选址总成本相比PSO算法减少11.4%,迭代次数相比减少28.6%,验证了IPSO算法可以得到质量更高的解,有效节省了选址成本。Aiming at the situation such as the continuous expansion of cold chain market demand and failure to rapidly upgrade the cold chain logistics industry,a location model of cold chain distribution center with the goal of cost minimization is built by terminal distribution enterprises for improving enterprises benfits.In order to solve the above problems,firstly,the Mean-shift clustering algorithm is used to process the demand points with unknown shapes and features,which ensures the rationality of distribution center location in spatial distribution and reduces the difficulty of operation.Secondly,the reverse search,mutation and intersection strategies are introduced to improve the traditional particle swarm optimization(PSO),and finally the improved particle swarm optimization(IPSO) is used to solve the model.The analysis results of the relevant cases show that the total cost of site selection using the IPSO and the number of iterations are 11.4% and 28.6% lower than that of the PSO respectively,which verifies that the IPSO can obtain a higher quality solution and effectively save the site selection cost.

关 键 词：物流配送中心选址 Mean-shift聚类算法 改进粒子群算法 成本最小化 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]