Abstract: The problem of time-based inventory service level in an emergency items supply system is studied in this paper. A two-echelon system consisting of a central supply location and multiple different local supply locations which faces with the stochastic items demands is considered. The local supply locations’ inventory strategy optimization model with time constraints under the transshipment mode based on the nearest principle is developed when the stockout happens in local supply locations, then, the effects to the inventory strategy with other transportation modes i.e. the stochastic transshipment mode and no transshipment mode based on the modification of the developed model are analyzed, and the PSO algorithm is used to solve and analyze the problem at last. The result shows that the inventory strategy with the nearest principle mode has a higher service level, and its strength is more significant compared to the others when the time constraints tightened.

Key words: emergency transshipment, time constraints, Particle Swarm Optimization（PSO）, the nearest principle, direct delivery

摘要： 主要研究有时限约束的应急供应系统的库存服务水平问题。考虑随机需求下的包含一个中央供应点和多个不同地方供应点的二级系统，当缺货发生时，基于就近原则的转运模式建立了有时限约束的地方供应点库存策略优化模型，在此基础上分析了随机转运和不转运两种模式对模型的影响，最后采用粒子群算法对模型进行求解和分析。结果表明：在有时限约束时，就近原则模式下的库存策略具有较高的服务水平，并且时限约束越紧，其优势越显著。

关键词: 应急转运, 时间限制, 粒子群算法, 就近原则, 直送