完全信息下的报童模型是研究单周期 下企业运营的核心方法, 制定不同情形下的订货决策是 提高其运作效率的主要策略. Scarf 在1958 年构建了需 求侧不确定时部分信息下的鲁棒报童模型, 经过~Gallego {\&} Moon 更简便的证明后Worst-case 的鲁棒订货方法得到了广泛的应用, 供应侧不确定时的订货问题由于求解的困难性导致运作管理学者长期受其困扰. 文章假设报童只知道供应商随机产出因子的均值和方差信息, 建立了Worst-case 的鲁棒订货模型, 给出了订货决策的闭式最优解. 研究表明: 鲁棒报童模型能够恰当的描述供应侧信息缺失下的企业订货行为, 鲁棒的订货策略能有效解决供应不确定的信息缺失环境下企业的订货决策问题.

Newsvendor model under full information is the core method to study business operations in single cycle. Formulating ordering strategies under different scenarios is the key tactics to improve its operational efficiency. Scarf has built up Robust newsvendor model with demand side uncertainty under partial information in 1958. Worst-case method which proved by Gallego {\&} Moon in a much easier way has been widely used. Due to the difficulty of solving, uncertain order problems on the supply side is a long-term obsession to scholars. In order to investigate enterprise's optimal order decision under supply uncertainty when supply information is not full, we assume that newsvendor retailer only know the mean and variance of supplier's random output. Then we build up Worst-case robust orders model and give out closed optimal solution to ordering decision. The result shows that Robust newsvendor model can describe the supply side enterprise ordering behavior without full information properly. Robust ordering strategy can effectively solve the problem of enterprise's ordering decision under the condition of supply uncertainty.