在接單後組裝生產環境中之整合式生產與配送排程問題

Abstract

面對日益激烈的競爭環境以及消費市場特性的改變，為了提高企業的競爭 力，愈來愈多企業選擇採用接單後生產(make-to-order, MTO)、接單後組裝 (assemble-to-order, STO) 或直接銷售(direct-sell)等營運模式，或採用即時化 (Just-in-time, JIT)的生產策略以減少存貨的積壓，這些營運模式特別適用於對時間 敏感度高、產品多樣性大、產品生命週期較短或產品銷售時期較短等特性的產 業。在接單後組裝的生產模式下，產品在組裝完畢後需在極短的時間內直接配送 至所需的顧客處，因此生產(production)與配送(distribution)兩階段作業間的協調性 成為營運是否有效率的關鍵。 近年來開始有少部份的研究由作業面(operational)探討適用於此種營運模式 的整合生產與配送問題，但大部分的研究重點都在於釐清問題的複雜度，並針對 一些受限制的特殊情況，發展出能於多項式時間(polynomial time)內找到最佳解的 演算法，而尚未發現有研究致力於發展能實際應用在解決較大規模問題的演算 法。本計畫將生產部份的作業以等效平行機台的排程模式描述，將配送部份的作 業描述為具有車容量限制的車輛途程問題，而以整合的方式考慮生產與配送兩階 段，在同時考量顧客服務水準與總配送成本的目標下，進行生產排程與配送路徑 規劃。本計畫將提出一個結合變數產生法(column generation)的分枝界限法(branch and bound)的方式，找尋此研究問題的最佳解，並將設計一個結合模擬退火法 (simulated annealing)的複合式禁忌搜尋法(tabu search)，在合理的運算時間內找尋 此研究問題的最佳近似解。本計畫將以測試問題評估所提出之演算法的求解品 質、效率與穩定性，並量化以整合方式探討生產與配送排程作業的效益。本計畫 之研究成果，能使得此種以整體觀點考慮生產與配送活動的模式更具實用性。

In current competitive global market, customers’ demand is highly unpredictable. Many enterprises have chosen to adopt make-to-order, assemble-to-order or direct-sell business model, or apply the just-in-time concept to lower their finish-good inventory. This type of business models is especially acceptable for those industries whose products are time-sensitive, with great variety, short life cycle or short selling period. In an assemble-to-order environment, products need to be delivered to their customers within a very short period of time right after being processed. The coordination between the production and distribution operations is very critical to determine the effectiveness of the business. Nowadays, a very limited number of studies start to consider the integrated production and distribution operations at individual job level from operational perspective. However, most of them are dedicated to complexity analysis or developing polynomial-time solvable algorithms for very restricted problems. It is rarely found any research to focus on solving problems in a larger scale. The proposed work discusses a more practical integrated production and distribution scheduling problem in which a parallel-machine scheduling model that incorporates routing decisions of unlimited numbers of delivery vehicles to serve customers at different locations is studied. The objective of the problem is to jointly develop a production and distribution schedule such that the cost considering both customer service level and distribution is minimized. A branch-and-bound algorithm based on column generation technique is proposed to solve this problem optimally. This study will also design a compound tabu search heuristic in which simulated annealing is used as an intensification strategy in order to find near optimal solutions in reasonable computation times. Computational analysis based on computer generated problems will be used to demonstrate the effectiveness of the proposed algorithms. Moreover, the integrated approach will be compared with the approach in which the production and distribution are considered separately and sequentially. It is expected that the research results would help make the study of integrated production and distribution problems more practical and with better application values.