考量離站距離影響需求變動的公共自行車站點選址問題

Abstract

近年來環保意識抬頭，各國政府皆積極地推動多項政策與措施，以達到環境保護、永續發展的目標。在交通方面，推廣綠色運具已成為一必然的趨勢，而本研究的焦點即在於近年在各城市快速發展的公共自行車系統(Public Bicycle System, PBS)。運輸系統的決策問題一般主要可分為「長程策略規劃」與「短程作業規劃」；前者常屬於基礎設施的結構性的設計，所牽涉之固定成本較高，完成後短期內不會再有大幅的更動；相對地，後者則在基礎設施固定的前提下，考量各種作業面的問題。本研究所探討之「公共自行車站點選址」問題，基本上為一為策略性、中長期的設計問題，考量站點的固定成本，以及向使用者收費所帶來的營收，以營運者長期系統利潤的最大化的目標，建構數學規劃模式以決定自行車站點的數目與設置位置。本研究探討公共自行車的使用需求與站點設置決策之間的相關性，將站點設置的決策、所衍生的站點和旅次起迄點之間的距離變動、以至最終所衍生之需求與營收的變動納入模式。本研究利用最佳化的技術(基因演算法等啟發式解法)，發展實務規模的大型問題的求解方法，並執行適當的數值測試，確認所發展模式與演算法的可適用性，根據本研究之數值例題之結果，可以提供公共自行車系統規畫者有效的決策輔助。

Over the past several decades, given the rising awareness about our environment, many governments in the world have adopted various policies to enhance environment protection and to facilitate sustainable development. For the transportation sector, promoting green transportation has been a well-recognized trend. This study focuses on the public bicycle system (PBS), which has been rapidly implemented in many cities. In general, the decision problems of transportation systems can be categorized as the long-term strategic decision and the short-term operational decision. As a typical type of decisions for the former, the design of the infrastructure usually involves a high capital investment and is changed within a short period of time. On the other hand, the latter may deal with the operational issues, given the fixed infrastructure. As a long-term strategic decision, the PBS station location problem in this study determines the number and locations of the stations to maximize the profit of the operator, by taking into account the fixed cost of the stations and the revenue collected from the bicycle users. In particular, given the inter-dependence between the PBS demand and the station location decision, the model considers the impact of the access distance on the PBS demand, subject to the station location decision. This study develops a genetic algorithm to handle the practical-scale problems. In order to validate the developed model and the solution algorithm, a suitable numerical experiment is performed. Based on the results, it is found that this study can provide appropriate decision support to PBS planners.