考虑时间序列关联的订单选择决策比较研究  

Comparative study of order selection decision considering time series association

作　　者：张人千[1] 

机构地区：[1]北京航空航天大学经济管理学院,北京100191

出　　处：《管理科学学报》2009年第3期44-55,共12页Journal of Management Sciences in China

基　　金：国家自然科学基金资助项目(70501002;70871005);航空科学基金资助项目(2007ZG51075)

摘　　要：在不同计划期的订单之间有可能存在时间序列关联,而现有生产计划模型大多不考虑这一关联关系对订单选择决策的影响.针对这一问题,以按单生产方式为背景,基于时间序列关联规则,建立了一个权衡当前与未来利润的订单选择决策模型,而将不考虑时序关联关系的订单选择模型作为比较研究的Benchmark问题.在数值仿真实验中,设计了一个动态随机的订单到达流,采用分阶段求解策略,以权重优先算法针对不同决策情景,分别基于两个模型进行决策并比较其决策结果.数据显示,在订单流中存在关联性的情况下,考虑时序关联关系的决策模型较优,证明在生产计划中考虑需求在时间方向上的关联性是必要的.Time series association exists in orders of different planning periods, however, the influences on order selection/acceptance of this association have not been considered in existing production planning models. To study this problem, the paper sets up an order selection decision model based on time series association rules. The model makes decisions after balancing the current and the future profit. As a benchmark for comparison, an order selection model that does not consider time series association of orders is built. In the numerical experiment, we design a dynamic and stochastic orders arrival process, and make decisions using the two models respectively. Under different scenarios, these two models are solved period by period based on a heuristic algorithm of weighted shortest processing time （WSPT）. The comparative results demonstrate that the model which considers time series association is better than the benchmark model if there is time series association in orders. So, it is necessary to take account order association influences in the production planning decision.

关 键 词：订单选择 时间序列关联 按单生产 

分 类 号：F203[经济管理—国民经济]