基于分层MTO订单的准入策略研究

Admission decision based on hierarchical MTO order

研究按订单生产(make to order,MTO)模式下随机订单的准入策略.本文提出的准入决策方法分为两步,首先在考虑随机订单的不同属性基础上,采用TOPSIS(technique for order preference by similarity to an ideal solution)方法计算新到订单与理想订单的接近程度,以此作为订单的综合收益,从而对新订单进行分层排序;其次采用MDP(Markov decision process)模型分别计算接受和拒绝订单的期望收益,从而决定是否接受该订单.研究发现,当市场需求接近或大于产能时,企业可从订单分层中获益;反之,市场需求显著小于产能时,分层不会带来额外收益.企业分层收益随需求增加或产能减少而递增,且与生产前置期呈倒U型关系.数值模拟发现,订单分层越精细,收益越大,且分层数约为需求与企业产能的比值时,订单分层收益接近最大值.

This paper studies order admission strategy for a Make to Order （MTO） production system with ran- dom arriving orders. Our method includes two steps. In the first step, the Technique for Order Preference by Similarity to an Ideal Solution （TOPSIS） is adopted to calculate the degree of closeness between new orders and ideal orders which will be regarded as the comprehensive income of the orders, which is used to further prioritize the new order. In the second step, whether to accept the new order or not is analyzed through Markov Decision Process （MDP） model. The result shows that when demand is close to or more than capacity, order hierarchy can increase benefits but not when the demand is largely less than the capacity. The income by strat- ifying orders will increase with the demand, decrease with the capacity, and has an inverted U-shaped rela- tionship with production lead time. The numerical result shows that the income is higher with more exquisite order hierarchy. When the hierarchy number is closer to the number of production lines, the income is closer to the maximum.