Abstract: To deal with the integrated decisions of location and inventory for the supply chain network design with multi-echelons, a new location-inventory model based on echelon stock policy is developed in which the selection of suppliers is considered. The objective of the model is to determine the number and locations of suppliers, plants and Distribution Centers（DCs）, the assignment of each customer to a DC, the transportation quantity between suppliers, plants and distribution centers, and the optimal order quantity in the opening plants and DCs, which minimizes the system-wide location, production, purchase, transportation, and inventory costs. The particle swarm optimization algorithm based on partial encoding is designed to solve the proposed model. The results of 20 numerical examples of different scale show that the model is effective and can be used to design some four echelons supply chain network, and the PSO algorithm designed performs better than mathematic optimization software of Lingo 9.0 whatever the precision or the speed, especially when there are many nodes in the supply chain.

Key words: location-inventory model, selection of suppliers, particle swarm optimization algorithm, partial encoding

摘要： 针对多级供应链网络设计中选址和库存一体化决策问题，基于梯级库存策略，建立了整合供应商选择的多层级选址-库存模型。模型以网络中供应商的选择成本、工厂和配送中心的打开成本、层级之间的运输成本、库存成本、采购成本和生产成本之和最小为目标，同时对供应商的选择、工厂和配送中心的选址、配送中心对顾客的分配、层级之间的运输量、工厂和配送中心的订货批量进行决策。为了求解所建立的模型，设计了基于部分编码的粒子群优化算法。20个不同规模的算例测试表明：所建立的模型是有效的，能用于多层级供应链网络的设计；所设计的算法无论是在求解精度，还是在运算速度上都明显优于数学优化软件Lingo 9.0，尤其是当供应链网络中总节点数较大时。

关键词: 选址-库存模型, 供应商选择, 粒子群优化算法, 部分编码