供應商主導多階存貨策略-以基因演算法求解

Vendor Managed Multi-echelon Inventory Policy Using a Genetic Algorithm

鍾生良(Shen-Lian Chung)

指導教授 ： 黃惠民

中原大學/工學院/工業工程研究所/博士(2004年)

https://doi.org/10.6840/cycu200400101

摘要

供應商存貨管理(VMI)是提升供應鏈管理效率中一種重要的手段，涉及供應商即時正確掌握下游零售商之銷售與存貨資訊，並按雙方協定的機制，適時逕行補充適量的貨品，此一策略已被廣泛地引用在不同的物業。本研究針對供應商主導多階存貨管理系統提出最佳補貨策略，以整合的方式來管理產品全面的原料供應、生產、配銷與零售，以最適的生產時間與配送批量來達到最低成本的存貨目標。然而多階整合存貨系統乃是複雜性問題，以傳統存貨管理的手法無法妥善解決。故本論文先以數學方法架構一多階整合存貨模式，再應用基因演算法求算，以穫得最佳解。本研究除著眼於非損耗性及損耗性物品之供應商多階整合補貨策略外，同時也考量超額存貨之回收問題；回收政策的良窳關乎著供應商和零售商合作的緊密與存貨成本的高低。實驗展示以數學分析演算結果，與以基因演算法求算結果比較，驗證基因演算法求解之品質。實例演算結果穫得，藉由供應商主導多階存貨策略能讓總存貨成本下降，但零售商存貨成本卻是上升，故要達成整合目標，供應商與零售商必須做到資訊互通與利潤分享的境界。
本論文主要內容如下：第一章主要介紹研究背景、目的、範圍、假設條件與論文組織。第二章為相關文獻探討與實驗評估。第三、四章探討非損耗性與損耗性兩階存貨模式並提出改良方法。第五、六章導出非損耗性與損耗性三階存貨模式。第七、八章發展出非損耗性與損耗性四階存貨模式，包括原料供應商、製造商、配銷商與零售商，並舉大型實例展示模式求解能力。另外第九、十章發展出非損耗性與損耗性買方超額存貨回收模式。最後，在第十一章提出本論文的結論與未來研究方向。

關鍵字

損耗性物品 ； 基因演算法 ； 供應鏈管理 ； 供應商存貨管理 ； 多階存貨

並列摘要

Integrated optimization via Vendor Managed Inventory (VMI) is an important approach for increasing supply chain efficiency. Having been widely used in various industries, VMI can achieve enterprise objectives through more accurate sales information and effective distribution of inventory in the supply chain. This study investigates an optimal ordering policy for a VMI system involving one vendor, multiple suppliers, multiple distributors and multiple retailers. Obtaining the optimal solution for multi-echelon cases is very difficult because of the complexity of the problems involving integrated VMI. Genetic Algorithm (GA), which is a search technique based on the mechanics of natural selection and natural genetics, is used to solve the global optimization problems. This study focuses on generalizing an integrated multi-echelon production inventory for non-deteriorating and deteriorating items to present an analytical model with genetic algorithm approach to optimize an integrated replenishment policy for the VMI systems. We also present an integrated excess stock return model to consider the views of both the vendor and the buyers. Returns policy is one of the important subjects related to the vendor and buyer coordination. It is commitment by vendor to retrieve excess stocks from buyer. An example of a multi-echelon inventory system is used to demonstrate the system. The solutions obtained by the genetic algorithm are compared with an independent decision by the economic order quantity. We can show that the integrated approach results in an impressive cost-reduction for the whole system.

並列關鍵字

Multi-echelon inventory ； Genetic algorithm ； Deteriorating items ； Supply chain management ； Vendor managed inventory

參考文獻

Bahari-Kashani, H., “Replenishment schedule for deteriorating items with time proportional demand,” Journal of the Operational Research Society, 40, 75-81 (1989).

Banerjee, S., A., “On ‘A quantity discount pricing model to increase vendor profits’,” Management Science, 32, 1513-1517 (1986).

Beltran, J. L. and D. Krass, “Dynamic lot sizing with returning items and disposals,” IIE Transactions, 34, 437-448 (2002).

Cachon, G. and M. Fisher, “Supply chain inventory management and the value of shared information,” Management Science, 46(8), 1032-1048 (2000).

Cetinkaya, S. and C. Y. Lee, “Stock replenishment and shipment scheduling for vendor-managed inventory systems,” Management Science, 46(2), 217-232 (2000).