建構交貨不確定下廠商評選與訂單配置之最佳化模式

Constructing A Portfolio Optimization Model for Vendor Selection and Order Allocation under Delivery Uncertainty

李家岩(Chia-Yen Lee)

指導教授 ： 簡禎富

國立清華大學/工學院/工業工程與工程管理學系/碩士(2006年)

摘要

近年來，隨著企業更加專注於核心競爭力的發展上，外包成為製造策略的重要一環，而專業的分工與外包可以使整體供應鏈的製造成本大幅下降並進一步減少生產風險、縮短生產週期時間與提高生產品質以達到公司長期獲利與永續經營終極目標。因此，廠商評選與訂單配置便因應時代所趨成為一常見且關鍵的決策問題。

廠商評選與訂單配置必需同時考量許多複雜因子及定性與定量的評估準則，為一高度複雜的半結構化問題。另一方面，在實際的產業環境中，外包的過程總是存在著諸多錯綜複雜的不確定性，這些原因都造成決策判斷更加困難。

本研究結合多準則決策分析與組合最佳化以建立交貨不確定下廠商評選與訂單配置之分析架構。其中，決策者可藉由廠商績效評估與風險分散的方法進而權衡績效、風險與成本三者目標間的關係來達到訂單之最適配置。本研究架構的步驟與方法包含架構問題、決策分析、資料彙整與預處理、屬性篩選、廠商績效評估、風險的鑑定與量化、模式建構與組合最佳化、輸出分析、建議與回饋。經由逐步的分析過程，最終結果可提供有效且穩健的廠商選擇與訂單配置以協助決策者之決策制定。此外，本研究以一數值案例說明及驗證所提出模式之效度。驗證結果顯示該模式所提出之訂單配置，能夠以較低之成本有效地提昇配置組合之整體績效並降低所承擔之整體風險。

關鍵字

廠商評選 ； 訂單配置 ； 外包 ； 產能規劃 ； 多準則決策分析 ； 組合最佳化 ； 不確定狀況下的決策問題 ； 穩健最佳化

並列摘要

Recently, outsourcing has played an important role in manufacturing strategy while enterprises have paid more attention to their own core competence. Particularly, specialized outsourcing not only lowers manufacturing costs of the overall supply chain but also further reduces production risk, shortens cycle time, and improve quality to ensure the ultimate objective of long-term profitability and sustainable operation. Thus, the vendor selection and order allocation (VSOA) becomes a well-known and critical problem.

The vendor selection and order allocation problem simultaneously considers many complicated factors and qualitative and quantitative attributes. It is indeed a highly complex semi-structured problem. Added to this, the outsourcing process in a real environment always exists uncertainty and vagueness which makes the decision-making more difficult.

This research incorporates multi-criteria decision analysis with portfolio optimization to construct a decision framework for the vendor selection and order allocation problem with performance and risk evaluation under delivery uncertainty. The objective of the proposed framework is to allocate appropriate orders to proper vendor and make effective tradeoff among the performance, risk and cost objectives. The methodologies of this research framework include decision analysis, feature selection, performance evaluation, risk identification and quantification, portfolio optimization, and simulation output analysis. Finally, a robust solution can be generated to assist in decision-making via step by step analysis. We conduct a numerical study to illustrate and validate the proposed framework. The result shows that suggested order allocation can take higher performance, minimize total cost, and achieve risk diversification.

並列關鍵字

Vendor Selection ； Order Allocation ； Outsourcing ； Capacity Planning ； Multi-Criteria Decision Making ； Portfolio Optimization ； Decision under Uncertainty ； Robust Optimization