Lagrangian Relaxation Method於三階層可修護最適備份件之應用

Finding Optimal Repairable Spare Parts for a Three-echelon Inventory System with The Application of Lagrangian Relaxation Method

官雅惠(Ya-Hei Kuan)

指導教授 ： 申生元博士

元智大學/資訊學院/資訊管理學系/碩士(2004年)

摘要

本研究的目的，係針對具有低需求、高價格特性且重要之可修護元件，發展出應用於國軍某後勤系統的存貨計算模式。由於高科技軍事武器系統及配備之採購費用昂貴，是以當相關元件功能故障或失效時，大都送由後勤維修單位進行修護。然因戰備任務需要，武器系統必須隨時處於正常可執行功能狀態，無法等待損壞元件經由長時間修護後方能運作（尤其對武器系統功能執行有重大影響之元件），是以相關重要元件之備份（Spare Parts）無法避免，因而規劃出一應用於國軍後勤武器系統之相關零附件備份模式。文中係針對單一組裝品項與多零附件備份之三階（場站、維修中心與基地）存貨模式進行規劃與探討，並期望能在有限的預算下，求解場站、維修中心與基地之最適備份量，使得各階層因缺貨所導致之總延遲時間（Backorder-Delay-Days）最小。
　　本研究首先分析維修中心與場站可再供應一正常功能組裝品項所需之時間，進而推導出基地組裝品項壞損後，可獲得一正常功能組裝品項之平均時間。並在給定預算經費下，構建以平均總延遲時間為目標之機率性整數規劃模式。由於此模式同時考量組裝品項與零附件之交互影響，因此提出以lagrangian relaxation為求解方法之兩階段式演算法。文末以相關研究資料進行分析與探討。

關鍵字

三階存貨模式 ； 可修復零附件 ； 延遲時間 ； Lagrangian Relaxation Method

並列摘要

The main objective of this research is to develop a decision tool in determining optimal stockings of each spare part for a three-echelon inventory system under limited budget constraint. This system we will model consists of a group of locations, called bases, maintenance centers, and a centeral depot. The items considered in the thesis are called recoverable items, that is, items can be repaired when they failed. Moreover, the items we mentioned consist of two categories: one is the assembly item, or module, and the other is the component. Each assembly is composed of many components. Briefly, the model is used to determine the best stock levels at each location for each item so as to achieve optimum inventory-system performance for a given or limited of investment, and is expecting to minimize the total backorder delay days which is caused by shortage.
We first compute the resupply time of a functionable item from the depot or maintenance centers, and then the resupply time to obtain a functionable item after a failure occurred at bases is calculated for the base-level. Accordingly, a probabilistic integer programming model is formulated with the side constraint of a given budget. In order to solve the model, a two-stage algorithm based on the Lagrangian Relaxation method is proposed. Finally, an example is used to illustrate the algorithm.

並列關鍵字

Three-echelon Inventory ； Recoverable Items ； Backorder-Delay-Days ； Lagrangian Relaxation Method

參考文獻

1.Angel Diaz*, “Modelling approaches to optimize spares in multi-echelon system”, International Journal of Logistics: Research and Applications, Vol.6, No.1-2, (2003) 51-62

3.Arthur F. Veinott, JR. and Harvey M. Wagner, “ Computing Optimal (s,S) Inventory Policies”, Management Science Vol.11, No.5, March 1965, 525-552

4.Books, R. and AM Geoffrion, "Finding Evertt’s Lagrange Multipliers By Linear Programming", Operations Research, 14:6, 1149-1153 (November — December 1966)

5.Clark, A., and Scarf, H., "Optimal policies for a multi-echelon inventory problem", Management Science 6 (1960) 475-490.

6.Cohen, M.A., Kamesam, P., Kleindorfer, P.R., Lee, H.L., and Tekerian, "Optimizer: IBM''s multi-echelon inventory system for managing service logistics", Interfaces, Vol.20 (Issue 1) (1990) 65-82.