冷链物流配送网络选址-路径-库存的优化

Location-Routing-Inventory Optimization for Cold-chain Logistics Distribution Network

为了控制冷链配送成本,优化3级冷链物流网络节点空间布局,提出了一种基于需求存在模糊随机性的选址-路径-库存多目标优化模型。以京津冀城市群为例,提供了优化方案。针对由物流基地、配送中心、末端网点构成的冷链物流3级配送网络,考虑到冷链产品存在订货提前期及需求具有模糊随机性,引入模糊随机变量,建立了CLRIP集成优化模型。首先,通过穷举法计算出各备选配送中心点的最佳订货周期和最优库存管理方式,将客户需求点与备选配送中心点进行随机组合,确定需要开放的备选配送中心点的位置与数量。然后,基于客户模糊随机需求,综合考虑土地、库存和配送等成本,进行选址-路径动态规划;利用改进禁忌搜索算法修正模型,直至达到最终要求,从而确定需要开放的备选配送中心点及其最优巡回路线。最后,对京津冀城市群多组实例进行分析。结果表明:使用模糊随机变量的试验数据中,最优解目标函数值和偏差稳定度均优于使用固定需求量的对照组数据,验证了CLRIP模型和改进算法的有效性。研究结果可为京津冀城市群冷链物流配送网络构建提供最优选址位置、配送路径和库存管理,为相关冷链物流配送网络优化提供实用可靠的参考方案。

To control the cold-chain distribution cost and optimize the spatial layout of three-tier cold-chain logistics distribution network,the multi-objective optimization model of location-routing-inventory based on fuzzy-stochastic demand was proposed.Taking Beijing-Tianjin-Hebei urban agglomeration as the study case,the optimal solution was provided.Considering order lead time and the fuzzy-stochastic nature of demand for cold-chain products,due to the three-tier cold-chain logistics distribution network composed of logistics base,distribution centers and terminal nodes,the fuzzy-stochastic variables were introduced to establish the integrated optimization model CLRIP.First,the optimal order cycle time and inventory management strategies for each candidate distribution center were determined by using the exhaustive method.The customers were randomly paired with the candidate distribution centers to determine the locations and quantities of distribution centers needed to open.Subsequently,based on the fuzzy-stochastic demand of customers and taking into account the costs(e.g.,land,inventory and distribution),the location-routing dynamic planning was carried out.The model was refined by using the improved Tabu search algorithm till the final requirement was achieved,thereby determining the distribution centers to be open and their optimal transportation routes.Finally,the multiple groups of experiments on Beijing-Tianjin-Hebei urban agglomeration were analyzed.The result indicates that the objective function value and deviation stability of the optimal solution were better than those of the control group with fixed demand,verifying the effectiveness of CLRIP model and the improved algorithm.The study findings provide optimal location,distribution routes,and inventory management strategies for constructing and optimizing cold-chain logistics distribution network for Beijing-Tianjin-Hebei urban agglomeration.