本論文旨在解決接單生產模式環境下，單一種產品之BOM內各階層各品項的生產規劃策略問題。所謂物料清單（Bill of Material, BOM），是指各個產品所有次裝配品、零件、原物料和所需要的數量之清單。問題中給定一BOM，若以一分割線 (Partition Line, PTL) 將整個BOM切割為二部份，PTL以下之品項以計劃式生產(Make-to-Stock, MTS) 的方式規劃，PTL以上之品項至完成品則以接單式生產(Make-to-Order, MTO) 的方式規劃。問題中具有MTO與MTS並存的生產環境之特性。 本研究延伸柯瓊琍 (2001) 研究，針對接單生產環境，並納入隨機需求、有限產能等因素，建立單一產品之BOM品項的生產決策問題(MTS/MTO)之非線性、整數規劃模型。目的為求得最小化成本下BOM 內各項生產規劃方式。與柯瓊琍 (2001) 不同的是，我們整理歸納出MTS與MTO之特性，並將此特性納入模型中。MTS品項通常有一存貨政策 (inventory policy)，MTO品項則不儲存任何存貨。 We study the planning strategy for each BOM item in the context of MTO environment. A BOM item can be MTO or MTS. We assume that the demand of end item follows a stochastic process. A production facility with limited capacity is approximated as an M/G/1 queue. The MTS/MTO decision for each BOM item is modeled as a zero-one variable. In this study, a MTS item follow an inventory (Q, r) policy. A MTO item does not keep inventory. We develop a nonlinear, integer programming model to decide each BOM item whether it is MTS or MTO. We restrict our approach within a single product format. We assume that the demand of end item follows a stochastic process. The facility with limited capacity is approximated as an M/G/1 queue. In particular, for MTS item, the production/inventory (Q, r) policy is set. For MTO item, no inventory is kept. And a committed lead-time is introduced. We extend the heuristic Ke (2001) developed and use AMPL/MINOS to solve item planning strategy and MTS inventory (Q, r) policy simultaneously. We consider a special case to implement the model. We also use a numerical test to evaluate the heuristic. Finally, a computational analysis is conducted to explore the impact of various problem parameters on the MTS/MTO decision. Several factors impact on the MTS/MTO decision and the (Q, r) policy, including unit holding cost, unit backorder cost, and set up time.
