備份件供應鏈系統運輸組態設計之研究

Abstract

本論文提出一降低備份件供應鏈系統營運成本的方法，此供應鏈系統為多個單位組成之網路，當其中任意兩個單位之間的運輸時間降低時，可減少備份件存貨水準，然而亦提高運輸費用。因此，本論文目標為設計一最佳運輸組態，所提方法主要分為四個階段：首先，本文使用可以有效設計備份件供應鏈系統之商用軟體OPUS10，作為計算存貨水準工具；其次為利用類神經網路方法建立快速運算機制，以估算OPUS10函數；然後發展一基因演算法搜尋近似最佳運輸組態；最後，發展模擬程式來精確估計此最佳運輸組態之績效。

This thesis proposes an approach to reduce the operating cost of a spare part supply chain system. Such a supply chain system is a network consisting of many nodes. The reduction of transportation time between any two nodes may reduce the inventory level of each node; at the price of paying more transportation fees. This research aims to identify an optimal transportation configuration. The proposed approach involves four major stages. First, we use OPUS10, a commercial software for effectively designing a spare part supply chain system, as a tool for computing the inventory levels for some sampled supply chain transportation configurations. Second, a neural network approach is used to establish a fast-computing mechanism to emulate the function of OPUS10. Third, a genetic algorithm is developed to identify a near optimal transportation configuration. Finally, a simulation program is developed to accurately evaluate the performance of the identified transportation configuration.