轉運航線之船期安排

Ship Scheduling Problem for Multiple Liner Services with Transferring Function

定期貨櫃航運服務為降低成本及提升效率，在航線設計上趨於網路化，其不僅提供近洋航線更多地區性點對點的直捷運送，亦加強遠洋航線接駁轉運的服務。因此對於航商而言，在各航線的船期安排內容中，必須充分發揮網路化的功能。本研究旨於建構一考慮轉運功能之船期安排模式，其中除考量個別航線之航程限制外，亦納入轉運港口之時間接續限制；在目標式方面，則以時窗偏好之概念，求取系統偏好加總值之最小化。在求解上，本研究建議一兩階段啟發式解法，以加快演算效率。經實例測試，模式求解結果不但能充分反應各港口之時窗偏好，亦能滿足航商在轉運方面的需求。此外，本研究發現轉運航線數過多，就時窗偏好的滿足而言，未必能為航商帶來更多的效益，因此，航商只需選取比較主要之航線，進行轉運船期安排，其他航線則以滿足其個別航線之偏好，將更為有利。亦即各航商於船期安排時，應以考慮各航線的市場為主，轉運為輔。

In order to offer better services and spend less costs, there is a trend that marine liners create a bundle of lines with liner service network characteristics. The containership routing design not only provides offshore sea routes with more direct shipping services from ports to ports, but also improves transshipment services for ocean routes. Thus, it is necessary for shipping carriers to take advantage of the function of the liner service network. The purpose of this research is to develop a model of transshipment cargo scheduling in which both the containership line and the time continuity between hub ports are taken into account. By using the concept of the preference for time slots at the ports of calls, the minimum preferred summation is obtained. This research uses a 2-stage heuristic search algorithm aimed at saving the CPU computational time. By testing the real cases, this model has been proven to be an effective tool in revealing the preference for time slots at the ports of calls, which is able to meet shipping carriers' transshipment requirements. In addition, this research finds that too many transferring routes will not bring shipping carriers effectiveness or efficiency. Therefore, it is favorable for shipping carriers to select primary containership lines in order to better schedule transshipment. The remaining containership lines will be arranged based on their preferences for time slots at the ports of calls. The primary concern for shipping carriers is the market of each containership line; the secondary concern then is transshipment.