基于提前经济指标的服务备件需求预测与库存管理

Service parts have been generating higher and higher value to the original equipment manufacturers, but the service parts inventory costs have been very expensive because of their high unit prices and high shortage costs. A few new features emerged from the data collected from a heavy machine equipment manufacturer about its service parts cost and demand information, this proposal suggests studying the roles of these new features in service parts inventory replenishment policy for minimizing the service parts inventory costs. The first new feature is that the service parts demands are typically not only dependent between periods, but also correlated with some economic indicators. Second, some service parts demands are intermittent, i.e., no demand occurs for a number of periods. Third, the upstream distribution centers of the service parts have to meet their own demands, in addition to replenishing downstream distribution centers. To characterize the impact of these new features, this project will investigate new demand forecasting methods to take advantage of historic information of both demands and leading economic indicators, study new inventory optimization theory to make use of the leading indicator information, construct new model to account for the intermittent demand, and develop new allocation scheme to serve demand in upstream distribution centers. Finally, this project will use the data collected from companies to validate these new methods and theories and to evaluate their economic value.

服务备件为原设备生产企业创造的价值越来越高，但是其库存管理的成本相当昂贵因为许多服务部件本身价格贵，而且缺货成本高。 基于重型机械设备生产企业提供的多类备件成本和多级配送网络数据，本项目将针对该数据呈现的新特点研究备件的补货策略，以最小化企业的库存成本。 第一个新特点是备件需求一般是非稳态的，未来备件需求不仅和其过去历史信息相关，而且和一些提前经济指标相关。第二个特点是一部分慢销备件的需求具有间歇性，即连续几期无需求。第三个特点是配送网络的上级节点不仅给下级节点供货，而且要满足其自身需求。针对这些特点，本项目将研究新预测方法以同时考虑备件历史需求信息和提前经济指标，新库存系统优化模型以利用提前经济指标提供的需求信息，新方法来刻画需求的间歇性，以及新动态库存分配策略应对上级节点需求。最后本项目将用企业实际数据校验提出的新方法和理论并评估其经济价值。

服务备件为原设备生产企业创造的价值越来越高，而其库存管理的成本也相当昂贵因为许多服务部件本身价格贵且其缺货成本非常高。 基于一家国际重型机械设备生产企业提供的多类备件成本和多级配送网络信息，本项目将针对该信息中的新特点研究备件的补货策略，以最小化企业的库存成本。本项目研究了服务备件管理四方面内容：结合提前经济指标和历史需求的备件需求预测、快销备件的动态库存管理、慢销备件的动态库存管理、多级节点的备件动态库存管理。 本研究提出和验证了新预测方法以同时考虑备件历史需求信息和提前经济指标，新库存系统优化模型以利用提前经济指标提供的需求信息，新方法来刻画需求的间歇性，以及在上级节点需求影响下的动态库存分配新方案。研究成果具有一定的实践价值。