單一產品及兩種顧客之MTS系統的生產控制與存貨分配問題

Abstract

本研究探討單一產品及兩種顧客需求的make-to-stock (MTS)生產系統的生產控制與存貨分配問題。生產控制問題是要決定該生產系統當存貨高達何種水準( S )時即該停止生產。 存貨分配問題是要決定當存貨低於何種水準( R )時，當低優先權顧客訂單到達時，需拒絕該訂單，將存貨保留等待優先權較高的顧客的訂單。傳統上這類問題是利用排隊理論及窮舉法，來決定最佳的S及R。本研究首先提出S最佳解上限的定理，隨之構建相關的經濟訂購量( Economic Order Quantity)模型，經由案例探討發現由相關的EOQ模型取得的S，可以當作是S最佳解的下限。利用改良式二切搜尋方法( Advanced Bisection Search )來搜尋S與R的最佳解以改善窮舉法的缺失。最後經由多個案例的實驗結果驗證我們所提方法的有效性。

In this study we consider the production control and inventory allocation problem of a make-to-stock (MTS) system with single product and two different demand classes. A production control policy specifies when to produce the product, that is, if the inventory rises to S level, then the MTS system stops to produce the product; otherwise, it is to produce. The inventory allocation problem is that if the inventory is above R level, then it is to satisfy all of two demand classes; otherwise, it is only to satisfy the higher priority demand class. Queueing method and Brute Force Search (BFS) method are generally used to determine the optimal levels of S and R. In this study we first provide a theorem to define the upper bound of S. Then we construct some related Economic Order Quantity (EOQ) models to define the lower bound of S. Then we propose an approach to identify the range of the optimal S and use the Advanced Bisection Search (ABS) method to find the optimal values. Finally, we verify experimentally the ABS method’s efficiency.