需求不确定条件下的制造商订单分配模型

制造商的订单分配作为供应链模型微观层面的重要组成部分,对提升整个供应链效率有很大影响,但需求层面的不确定因素加大了订单分配的难度。以按比例分配为原则,讨论在需求不确定条件下完全信息与不完全信息两类多供应商-单制造商的订单分配模型。重点研究完全信息条件下各方的分散决策和集中决策,由于后者能避免各参与方对其他决策方的边际影响,所以能够实现供应链总利润的最大化;其次又将不完全信息引入模型,讨论制造商如何通过折算因子结合已有信息对供应商的私人信息进行估计,进而做出决策。最后以需求服从正态分布为例对两类模型进行验证。

Manufacturer's Order Allocation Model under Demand Uncertainty

As an important part of the supply chain model at the micro level, manufacturer's order allocation has a great influence on improving the efficiency of the supply chain. In the meanwhile, demand uncertainty, such as the policy cycle shortening and consumption patterns changing, increases the difficulty of the manufacturer's order allocation. However, supply chain management disruption caused by demand uncertainty has not been generally studied in the current literature. In order to maximize their profits or total supply chain profits under certain allocation rule, how supply chain participants should make decisions, which is discussed especially in this paper. Some positive explorations are made in order allocation models, which demand uncertainty is considered and incomplete information is introduced into the supply chain participants. Two kinds of the multi-suppliers and one-manufacturer models are analyzed within the method of proportional distribution under demand uncertainty. A supply chain model is emphasized on the study of decentralized decision and centralized decision based on complete information, which concludes that the latter can maximize the interest of supply chain because of the avoidance of marginal effect caused by one participant to another when making decisions. And then, another model is established upon incomplete information, which pays emphasis on the decision-making mechanism of manufacture based on its estimate of suppliers' private information with discounting factor and existing information. In addition, the risk tolerance of suppliers depends mainly on their production capacities and variable costs, and the intention of supplier getting order is proportional to the risk tolerance, inversely to the unit punishment. Finally, the aforementioned two models are validated by an example of demand obeying normal distribution, given that the demand of most goods is normal distribution in real life.