- 學位論文
Demand-Driven Disassembly Plan for a Robust Closed-Loop Supply Chain System
穩健閉迴路供應鏈中的需求導向拆解規劃
摘要
In the wake of imminent government regulations and consumer awareness of environment-friendly manufacturing, the manufacturers must take the responsibility of the used products. Closed-loop supply chain system, which integrates the forward and reverse logistics, is a desirable policy for retaining recoverable resources and extending life cycle of products. In this thesis, we propose a mixed-integer programming model to contend a disassembly planning problem under a closed-loop supply chain system with multi-period, multiple products, and hierarchical product’s structure. The objective of the model is to determine the optimal volume and timing of each type of end-of-life (EOL) products to be recycled from end-users. The recycled products are then disassembled to be reused, remanufactured, repaired, or disposed. The optimal disassembly and recovery strategies are also determined under the constraints of capacities and satisfying demand for products. Furthermore, the proposed model accounts for the market mechanism under a closed-loop environment, including timing, quality and quantity issues of recycled EOL products. The numerical results of the illustrative example show the validity of the model being able to provide the required information for policy making. However, due to the uncertainty exists in the proposed closed-loop supply chain system, a series of scenario analysis is conducted to investigate the sensitivities of periodic demands, quality of recycled product, and timing of product return, and also the resulted impacts on optimal strategy. The results suggest that all of these factors are critical and substantial to influence the decision. Therefore, to mitigate the difficulty of decision-making resulted from the uncertainty in these factors, we propose a two-stage robust programming approach to determine a robust solution that provides the most adequate strategy by considering future scenarios at the beginning with a decision-maker’s attitude towards risk. The first stage decision is to determine a compromised solution that is close to optimal solution for every scenario while remaining a certain level of infeasibility of constraints, such as unsatisfied demand. Afterward, when the outcome of scenario realizes, the second stage decision, for example, inventory volume, is conducted to become a buffer for absorbing or mitigating uncertainty impacts. Furthermore, the computational results confirmed the trade-off relationship between solution robustness and model robustness, which are also kernel results of the robust model apart from expected profit. Finally, a contingency plan of a robust decision is developed, providing higher profit when shortage and excessive of procurement are both allowed.
並列摘要
現今隨著政府環境法規的要求以及消費者環境意識的興起，產品製造商必須負起將產品回收的責任。閉迴路供應鏈系統是在此議題下的一個值得採用的策略。它整合了正向及逆向物流，藉由回收再利用，保留可再使用的資源，並且延長產品的生命週期。在本論文中，我們針對一個閉迴路供應鏈系統提出一個求解需求導向拆解規劃問題的混整數規劃模型，同時考慮多期、多產品、以及產品的階層結構。模型的目的是決定從最終使用者手中各種壽終產品的回收時間及數量。回收產品進而被拆解以做為再使用、再製造、維修、與丟棄等再利用方式。同時，模型也決定這些回收產品的最佳拆解及再利用策略，且滿足產能與產品需求的限制。此外，本論文所提出的模型考量了在閉迴路環境中的一些市場機制，例如回收產品時間、品質、及數量的議題。經過數據測試讓模型獲得驗證，且所提供的結果具有效性，提供決策者足夠的資訊。 然而，由於在本研究所提出的閉迴路架構中，產品需求、回收產品品質、以及回收時間存在著許多不確定性，因此我們進行一系列的情境分析來確認這些不確定性因素的敏感度以及它們對於最佳策略的影響。而結果顯示，其中所有的參數皆顯著地改變了決策。因此，為了減緩不確定性造成的決策困難，本論文一開始即考慮可能的未來情境，在決策者對於風險的偏好態度下，以二階段穩健規劃方法來決定一個最適當的穩健決策。第一階段目的在於找出一個折衷的決策，以期能夠在所有情境中接近最佳解，同時允許一定程度的不可行性，例如需求的不滿足。而當情境確定發生的當下，決策者選用適合此情境的第二階段決策，並成為一個緩衝機制以吸收或減低不確定性所造成的影響。另外，測試結果也顯現出解穩健性與模型穩健性中的權衡關係。最後，當超額採購與緊急採購允許的情況下，本論文提供了一延伸的模型以求解應變計畫下的穩健決策，其利潤也獲得了改進。