供需不确定下制造企业订单承诺决策研究

With the development of information technology and increasing competition, the uncertain supply and demand bring great opportunity and challenge for order promising of the manufacturing companies. It becomes very urgent for researchers and practitioners to solve the problem on how to respond quickly and effectively to requests for quotation from customers with limited resources to obtain maximum profit. By applying revenue management method, this research is going to study the order promising problem of manufacturing companies under multiple types of resource constraints and supply&demand uncertainties associated with the possibility of order cancellation. This project is going to propose a stochastic dynamic programming model and a simulation&optimization model to solve it. The following important problems are going to be tackled to provide scientific basis for order promising in manufacturing industry: (1) analyzing the factors that influence the promising policies from the perspective of characteristic of order and system parameters; (2) analyzing the influence of order cancellation to optimal policy and profit; (3) construction of the overbooking or penalty mechanism to reduce the risk caused by order cancellation. Finally, empirical research is conducted to demonstrate the effectiveness of the proposed theory and method as well as its application prospects.

信息技术的进步和市场竞争的加剧增加了制造业供给和需求的不确定性，给企业的订单承诺管理带来了极大的机遇和挑战。如何在供需不确定性的现实情况下，及时、准确地对客户订单作出承诺，并有效地分配企业有限的资源，实现利润的最大化，是理论者和实践者面临的一个迫切需要解决的现实问题。本项目拟应用收益管理的思想和方法，研究多种资源约束及供需不确定下（包括顾客取消订单的可能性）制造企业的订单承诺问题，通过构建随机动态规划模型，并利用仿真优化技术建立系统决策模型，对问题进行优化求解。本项目研究预期通过解决如下三个重要问题，为制造业的订单承诺决策提供科学依据：（1）从订单属性和系统结构参数两个维度，揭示影响订单承诺决策的相关因素；（2）量化分析客户取消订单对企业决策和利润的影响；（3）构建有效的超订或惩罚机制，降低客户取消订单对企业带来风险。最后，通过实证研究验证本项目理论和方法的有效性，展示其应用前景。

在实际运营中，快速而可靠的订单承诺决策是制造企业必须解决的一大难题，尤其当短期内原材料供应不足或产能不足等原因导致产品供不应求时，在考虑到未来高利润订单的机会成本时，如何选择接受哪些订单对于企业就显得尤为重要。本项目采用仿真优化集成方法对不确定环境下订单承诺问题展开研究。通过建立随机规划模型以及数学分析，发现存在最优的订单接受策略，通过仿真分析对比了不同订单承诺策略的优劣以及不确定性及系统参数对绩效的影响。研究表明：（1）订单履行过程中的不确定性会降低企业的收益；（2）将顾客分等级并利用收益管理的思想可以在实际运营中提高企业收益；（3）每期对每类顾客存在一个阈值，当供应大于该阈值时则接收当前订单，否则拒绝；（4）允许提前预定比只考虑当前库存更能提高企业收益；（5）在有碳排放约束情况下，碳排放权交易市场可以鼓励企业生产绿色产品，减少排放，而且更高的碳价可以促使企业投资绿色技术。