請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/83478
標題: | 基於組合最佳化下的自適應探索式供貨模型 Adaptive Exploration-based Allocation Model under Bundle Optimization |
作者: | 丁立恆 Li-Heng Ting |
指導教授: | 楊曙榮 Shu Jung Sunny Yang |
關鍵字: | 品質控制,捆束最佳化,自適應配對, Quality control,Bundle optimization,Adaptive matching, |
出版年 : | 2022 |
學位: | 碩士 |
摘要: | 對製造業而言,品質管控問題長久以來都是非常重要的問題,因為就算再先進的工廠都會產出次級品或瑕疵品。與眾多的研究相同,為了避免沒有足夠良品可以供貨的窘境,通常廠商會進行超額生產或透過投資生產技術來提升生產技術。然而,每個顧客的偏好其實各不相同,對一個顧客來說是次級品或瑕疵品的產品對其他顧客可能是良品。此論文對廠商的兩階段決策問題進行建模,並在客戶偏好未完全公開且有生產不確定性的情況進行研究,透過在組合最佳化下的可運算的適應性探索式兩階段模型來解決良品與次級品的分配問題。在實務上,公司的存貨分配問題通常是兩個部門負責掌管,分別是業務(顧客管理部門)與存貨管理部門。組合最佳化模型透過保存客戶更喜歡的存貨來避免良品斷貨問題,與此同時,探索式供貨模型透過對客戶進行溫和的偏好測試來探索客戶的偏好並避免客訴危機。而此論文將此兩階段模型應用於來自亞洲的大型塑膠生產商之一的生產與供貨資料,我們發現此兩階段模型不但可以避免25% 的退貨案件,也可以幫助公司更了解顧客的偏好。 Quality control problem is important to manufacturing section since seven the most advanced factory will produce defectives or secondary products. In case of quality disruption, most of the manufacturers will over produce or investing in production technology, which is identical to the focus of most researches. However, every customer has their own taste. The secondary to a customer might be normal product to the others. This paper develops an adaptively explorative tractable methodology for the allocation problem under bundle optimization with production and preference uncertainty, which delineates the two-stage decision- making process in a company. In practice, inventory control and allocating inventory to customers individually are two seperated responsiblilities for two departments, sales (customer management department) and inventory control department. The explorative allocation model helps sales to mildly explore customers' preferences without causing customer management crisis, while the bundle optimization improves the inventory management by preserving relatively valuable profiles to avoid quality disruption. Using the data from one of the biggest plastic manufacturer in Asia, we show our model can potentially reduce 25% of the return cases for the company while knowing customers' preferences better. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/83478 |
DOI: | 10.6342/NTU202203065 |
全文授權: | 未授權 |
顯示於系所單位: | 統計碩士學位學程 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-110-2.pdf 目前未授權公開取用 | 1.86 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。