應用蟻群演算法於求解供應鏈中整合產品製造與成品配送兩階段問題

Abstract

產品製造與成品配送是供應鏈中相當重要的兩個階段，由於顧客需求的快速變動以及產品的高度客製化，現今許多企業採用接單式生產(make to order)或直接銷售(direct order)模式而被迫減少存貨，但同時仍必須快速回應顧客的要求以維持其競爭力。此種現象增加了供應鏈中產品製造與成品配送作業的互動，並提昇了整合這兩階段研究的實用性。 本研究探討整合供應鏈中產品製造與成品配送兩階段問題，在這種問題中，訂單會先在一組非等效平行機台上(unrelated parallel machine)進行加工，加工完畢後則由具容量限制的車輛配送至顧客的所在地。訂單的完工時間定義為送達顧客的時間，而問題的目標則是求出一個整合製造與配送的排程，以最小化加權後的總完工時間。 然而整合製造與配送兩階段問題的複雜度為未定多項式難度(NP-hard)，為了在合理的時間內求得品質良好的近似最佳解，本研究採用蟻群演算法(ant colony optimization)來求解此類問題，並透過電腦實驗來測試演算法的求解品質及穩健性，期望本研究結果能有助於提昇整合製造與配送兩階段問題的實用性及應用價值。

Due to the rapid-changing market demand and highly customized product requirements, many enterprises have adopted make-to-order or direct-order business model. In this type of business model, enterprises are forced to lower the amount of inventory needed across their supply chain but still have to be more responsive to customers’ requirements. Reduced inventory creates a closer interaction between production and distribution activities and thus increases the practical usefulness of integrated models. We consider an integrated production and distribution problem at the individual job level in this study. In this problem, jobs are first processed by one of a set of unrelated parallel machines and then distributed by vehicles with limited capacity to the corresponding customer locations. The completion time of a job is defined as the time when it is delivered to its customer. The objective is to find a joint production and distribution schedule so that the total weighted completion time is minimized. The complexity of the studied problem is NP-hard. Therefore, we used an ant colony algorithm to solve this problem in order to find near-optimal solutions in reasonable computation times. Computational analysis is performed to evaluate the effectiveness and stability of the proposed approach. We expect our research results to help make the study of integrated production and distribution problems more practical and with better application values.