在近代生產管理中，日本企業界提出了引起世人矚目的剛好及時 (Just-In-Time，JIT) 生產系統。所謂「剛好及時」即是指將需要的物品在需要的時間供應符合要求的品質與必須的數量到達需要的地點。其成功之道乃在於透過多方面持續不斷的努力與改進，將資源的利用最佳化，進而以最經濟的方式生產高品質的產品，使企業體獲得對外的競爭優勢。欲達成JIT的目標，投資資金以降低設置成本、縮短前置時間、改善產品與製程的品質水準等活動，被認為是可行且有效的方法；換言之，傳統存貨策略中所經常假設的固定且不可控制的參數，如設置成本、前置時間等事實上是可以控制的。環顧現有的存貨管理文獻，已有諸多的學者建構了符合JIT精神的存貨計量模型，此類型的研究大多分別以降低設置成本、縮短前置時間，或改善品質水準為主題。然而，少有研究是同時控制其中兩個或多個可變動的參數，這將使得系統中的浪費無法減至最低，也因此無法節省較多的成本。有鑑於此，本文試圖融合以往學者所提出的概念，以循序漸進方式逐一放寬傳統存貨系統中固定參數值的假設，建構更為完善、使用範疇更為廣泛的存貨模型。
本論文係以JIT的目標為導向，針對連續性檢查的批量請購點 (Q,r) 存貨系統，提出八個機率性存貨數學模型。每一個模型均允許缺貨發生且在缺貨期間考慮欠撥與銷售損失混合的情形，共同的決策變數有訂購數量、請購點與前置時間。在第二章中我們探討以投資資金來降低設置成本的問題，模型中的設置成本為決策變數之一。第三章則進一步考慮品質與批量的互動關係，並討論品質的改善問題；品質水準將被視為可控制的因子，為決策變數之一。第四章則延續第二、三章的研究，建構可同時控制訂購數量、請購點、設置成本、品質水準與前置時間的模型。第五章，我們將傳統目標函數中的缺貨成本項改以服務水準限制式來取代，並繼續討論第四章的主題投資資金以改善生產過程之品質、降低設置成本與縮短前置時間。每一章，我們均提出二個模型，一為前置時間內需求量的機率分配服從常態的情形，並建立求問題最適解的演算法；另一為前置時間內需求量的機率分配型式為未知，而僅已知其平均數與標準差的情形，並運用分配不拘大中取小準則來求得問題的最適解。對於所提出的每一個模型，我們均以數值範例說明投資資金去改變問題參數值時對存貨系統的影響及其所帶來的利益。

Among the modern production management, the Just-In-Time (JIT) production system that developed by the Japanese business organization has received a great deal of attention from many practitioners and academicians all over the world. ‘Just-In-Time’ means that it has the right quality and right quantity in the right place at the right time. The successful JIT is through the various efforts and continuous improvements that make the efficient usage of resources and produce the high quality products in the most economical manner, so as to gain the competitive advantages for the business enterprise. Many activities, such as reducing the setup cost, shortening the lead time, and improving the quality of production processes and products, are recognized as the feasible and effective ways to achieve the goal of JIT. In other viewpoints, the factors (setup cost, lead time, and quality) mentioned above are often assumed as fixed constant and uncontrollable in the traditional inventory models, but are controllable in practice. Viewing the current literature in the area of inventory management, it can be found there are numbers of quantitative inventory models associated with the JIT spirits have been proposed. Most of the earlier researchers are concentrated on investigating one of the topics  setup cost reduction, lead time reduction, or quality improvement. However, little work has been done on simultaneously controlling two or more changeable parameters, and hence, the waste in using the current systems can not be reduced to the minimum level, so does the savings on larger cost can not be achieved. For these reasons, this thesis attempts to involve the concepts of previous scholars, and seeks to formulate the more appropriate and extensive inventory models by gradually relaxing the assumptions of fixed parameters in the traditional inventory models.
Oriented on the goal of JIT, this thesis proposes eight stochastic inventory mathematical models for the continuous review (Q,r) inventory systems. In all models, the shortages are allowed and the total amount of stockout during the stockout period is considered to be a mixture of backorders and lost sales. In addition, the common decision variables are order quantity, reorder point, and lead time. In chapter 2, we discuss the problem of investing capital in reducing setup cost, where the setup cost is treated as one of the decision variables in the model. In chapter 3, we further consider the possible relationship between quality and lot size, and investigate the quality improvement problem in which the quality level is viewed as a controllable factor and is one of the decision variables. In chapter 4, we continue the studies of chapters 2 and 3, and propose a model to control the order quantity, reorder point, quality level, setup cost, and lead time, simultaneously. In chapter 5, we employ a service level constraint to replace the shortage cost in the objective function, and discuss the same problem as in chapter 4. Furthermore, two models are formulated and solved in each chapter. First, we consider the case where the demand during lead time follows a normal distribution and establish an algorithm of finding the optimal solution. Second, we consider the case where the distributional form of lead time demand is unknown but merely the mean and standard deviation are known; we apply the minimax distribution free approach to solve the optimal solution. For all models proposed in this thesis, we utilize the numerical examples to illustrate the effects of inventory systems associated with investing capital in changing the values of parameters.
頁次
表目錄..……………………………………………………….……四
圖目錄………………………………………………….…………..五
使用符號一覽表….……………………………….…………... 六
基本假設..……………………………………………………… 八
第一章 緒論 ……………………………………….……………1
1.1 研究動機與目的 ……………………………………… 1
1.2 相關文獻探討 ………………………………………… 3
1.3 本文結構 ……………………………………………… 8
第二章 降低設置成本與縮短前置時間的連續性檢查
存貨模型.………………………………….…………. 10
2.1 前言………………………………………………… 10
2.2 符號說明與假設…………………………………... 12
2.3 基本模型……………………………………….…... 13
2.4 需求量之機率分配為常態時的模型..…………….… 16
2.5 需求量之機率分配型式未知時的模型.………… 22
2.6 範例…………………………………………………… 25
第三章 在生產過程不完備的連續性檢查存貨模型中
改善生產過程之品質與縮短前置時間…………….. 30
3.1 前言………………………………………………… 30
3.2 符號說明與假設…………………………………… 32
3.3 基本模型………………………….………………… 33
3.4 需求量之機率分配為常態時的模型………………… 35
3.5 需求量之機率分配型式未知時的模型…………. 39
3.6 範例…………………………………………………… 42
第四章 改善品質、降低設置成本與縮短前置時間的
連續性檢查存貨模型………………………………. 45
4.1 前言………………………………………………… 45
4.2 基本模型……………………………………………. 46
4.3 需求量之機率分配為常態時的模型………………… 47
4.4 需求量之機率分配型式未知時的模型…………. 53
4.5 範例…………………………………………………… 55
第五章 在含有服務水準限制式的連續性檢查存貨模
型中控制可變動的參數值………………………….. 60
5.1 前言………………………………………………… 60
5.2 基本模型……………………………………………. 61
5.3 需求量之機率分配為常態時的模型………………… 63
5.4 需求量之機率分配型式未知時的模型…………. 69
5.5 範例…………………………………………………… 76
第六章 結論…………………………………………………. 79
6.1 主要研究成果……………………………………… 79
6.2 未來研究方向……………………………………….. 82
參考文獻……………………………………………………….. 84
附錄A ………………….…………………………………..… 90
附錄B………………….………………………………...…. 93
附錄C ………………….…………………………………..… 96
附錄D………………….………………………………...…. 100
表 目 錄
表號 頁次
表 2.1 前置時間內各成份的相關資料……………… 26
表 2.2 例題一的最適策略彙整表………………………….27
表 2.3 例題二的最適策略彙整表………………………….28
表 3.1 例題三的最適策略彙整表………………………….43
表 3.2 例題四的最適策略彙整表………………………….44
表 4.1 例題五的最適策略彙整表………………………….56
表 4.2 例題六的最適策略彙整表………………………….58
表 5.1 例題七的最適策略彙整表………………………….77
表 5.2 例題八的最適策略彙整表…………………….…..78
圖 目 錄
圖號 頁次
圖 1.1 本研究結構流程圖……………………………… 9
圖 A-1 ， 及 圖………………………………. 92
圖 B-1 圖……………………………………………… 95