貨櫃航線網路設計模式之研究

A Network Design Model for the Containership Routing Problem

定期貨櫃航運為現今海運市場之主流，為降低營運成本及提升效率，貨櫃航線設計問題遂成為航商需面對之重要課題。貨櫃航線之設計乃基於航運市場之考量，設計貨櫃船隊之繞行路線，然隨著貨櫃航線之聯營及網路化，使航線設計問題日趨複雜，實有賴一分析工具協助規劃設計。本文旨在考量航線設計之基本實務因素（如往返繞行、航段容量限制等），以及航運之發展趨勢（如聯營化、網路化等）等因素，據以構建一貨櫃航線網路設計模式，以期有效協助航商進行航線設計作業。本文乃利用網路流量之觀念來探討航線設計問題。經分析，該設計問題必須同時考量多個不同的貨櫃起迄對（貨櫃流），以及貨櫃船舶（船舶流）在航運網路上的流動，故本質上係屬一多元商品網路設計問題。在模式構建與測試上，本文所構建之航線網路設計模式可同時考量主、接駁航線之設計、及航線聯營因素，經數個簡例測試，顯示模式之運算結果尚稱合理。 Nowadays containership liner service is the major way to transport containers by sea. In order to offer better service and spend less cost, shipping carriers are facing the containership routing problem. This problem has become more complicated because shipping carriers not only design one containership line, but also bundle of lines that construct a liner service network. For instance, a major shipping carrier may have seven containership lines for the trans-pacific service. Furthermore, containership lines can be operated by shipping carriers themselves as well as with help of consortium, leasing or exchanging container slots from other carriers. This paper makes an attempt to develop an optimization model to assist shipping carriers to solve this containership routing problem. The optimization model is formulated as a multicommodity network design problem, which takes ship flows and container flows into account. That is, if containership lines are designed, ships will periodically visit en route ports in their lines, and containers can be unloaded or loaded in these ports and transported from their origins to their destinations. Furthermore, designed containership lines are divided into oceangoing ones and feeder ones. This optimization model is tested by some test cases. The results seem promising. This model is expected to help shipping carriers to solve the containership routing problem.

Nowadays containership liner service is the major way to transport containers by sea. In order to offer better service and spend less cost, shipping carriers are facing the containership routing problem. This problem has become more complicated because shipping carriers not only design one containership line, but also bundle of lines that construct a liner service network. For instance, a major shipping carrier may have seven containership lines for the trans-pacific service. Furthermore, containership lines can be operated by shipping carriers themselves as well as with help of consortium, leasing or exchanging container slots from other carriers. This paper makes an attempt to develop an optimization model to assist shipping carriers to solve this containership routing problem. The optimization model is formulated as a multicommodity network design problem, which takes ship flows and container flows into account. That is, if containership lines are designed, ships will periodically visit en route ports in their lines, and containers can be unloaded or loaded in these ports and transported from their origins to their destinations. Furthermore, designed containership lines are divided into oceangoing ones and feeder ones. This optimization model is tested by some test cases. The results seem promising. This model is expected to help shipping carriers to solve the containership routing problem.