Abstract: A supply chain consisting of a manufacturer (buyer), a supplier, and a customer faces the logistics path closure or congestion risks brought about by large activities in metropolis. The manufacturer runs in the policy of ‘order-up-to-level’ inventory under quantity uncertainty, customer order uncertain, and subsided by the government. A multi horizon Markov state updating model involves in the expected cost optimization when there is a state transfer between the state of logistics path closure or congestion and the normal state. An approximate optimal solution may be found in the optimal ‘order-up-to-level’ inventory. It is found that the best ‘order-up-to-level’ inventory is more affected by the mean of congestion quantity and customer order service level. It is also found that the ‘order-up-to-level’ inventory is little affected by government subsidies but can decrease expected inventory cost. The best way to reduce the risk of logistics path closure or congestion is to increase the order quantity.

Key words: logistics path closure, logistics path congestion, inventory control strategy

摘要： 大型城市活动或重大自然灾害引起的物流路径关闭和拥塞风险问题受到广泛关注。通过建立一个多期马尔科夫链状态更新系统模型，分析物流供货系统在路径关闭和拥塞两种情形下的最优库存控制策略问题。结果显示：最优基本库存水平随着用户订单服务水平（报童分位数值）、关闭中断概率以及拥塞数量均值的增加而增大，但是增加显著水平存在差异。仿真研究了厂商如何提高订货量来减轻物流路径关闭或拥塞风险。这对于科学地应对供应链关闭和拥塞风险具有一定的理论和现实意义。

关键词: 物流路径关闭, 物流路径拥塞, 库存控制策略