Abstract: Inventory and transportation are the topmost main activities in modern logistics，and how to optimize these two activities is a vitally challenging work to the vendor-managed-inventory mode.Taking the two-echelon system consisted of one distribution center and many retailers as the object，and aiming to minimize the mean total cost in the system planning horizon，the paper analyzes all of the cost，including the retailer’s stocking，shortage penalty，and distribution center’s replenishment，holding，distribution，and models for the Inventory-Routing Problem（IRP）.Further，according to the retailer’s demand stochastic property，the bound of its stocking is estimated.A primary ordering policy and a distributing priority principle are selected for the distribution center based on the two-stage characteristic of the IRP.Borrowing idea from solving famous traveling-salesman problem，a heuristic algorithm is proposed for the IRP.As a result，a simulating problem，which is solved by computer program with Matlab7.0，is used to test the effectiveness of the optimization method and the algorithm，the IRP outcome saves cost about 7.3%，compared with their separate optimization.

Key words: logistics and supply chain, inventory-routing problem, distribution, stochastic demand, heuristic algorithm

摘要： 库存和运输是物流活动的最主要环节，如何同时对二者进行优化，是供应商管理库存必须解决的核心问题。针对一个配送中心为多个零售商实施统一配送的两级系统，以最小化系统计划期平均总费用为目标，分析包括零售商存储费、缺货费，以及配送中心进货费、存储费和配送费的系统全部费用组成，整合库存与配送建立了库存路径问题数学模型。根据零售商需求的随机性和库存路径问题本身的两阶段性，估计零售商库存上下限，选择配送中心订货策略及配送优先原则，借鉴旅行商问题求解思路，设计了求解问题的启发式算法。用Matlab7.0编程实现仿真算例求解表明，整合优化比分别优化节省平均总费用6.2%。

关键词: 物流与供应链, 库存路径问题, 配送, 随机需求, 启发式算法