不同選擇結構下之城際與接駁運具選擇模式

Abstract

近年來，由於經濟發展及產業結構改變，汽機車持有與使用數量快速成長，人口也快速且大量地朝都市地區集中，使得有限的道路空間因無法負荷龐大的私人運具使用需求而面臨擁擠的情況，而私人運具的使用也同時帶來空氣與噪音等污染，故鼓勵公共運輸降低民眾對私人運具使用的依賴性是目前世界各國正積極推展的重要交通政策。然而，旅運者使用公共運輸的意願與其使用需求息息相關，故在訂定各項交通政策時必須先瞭解其旅運行為，才能提供符合需求的政策作為，進而達成提高公共運輸使用率的目標。 爰此，本研究旨在探討影響城際運輸運具選擇的關鍵因素，並據以提出能有效提升公共運輸使用率之對策。個體離散選擇模式為本篇研究主要的研究理論與方法，並透過大規模全國性的問卷調查，來蒐集臺灣地區旅運者的選擇偏好及其社經背景等資料。由於不同區域的旅運者(例如：偏遠郊區與市中心區)在運具的選擇行為上可能會有不同的決策過程，為能反映出不同區域特性及交通條件之下旅運者選擇行為的差異性，本研究將樣本分為3個區域：偏遠地區、都市區及市中心區，並利用不同模式巢層架構之設定：架構一係上巢層為城際運具選擇與下巢層為接駁運具選擇、架構二係上巢層為接駁運具選擇與下巢層為城際運具選擇，分別進行運具選擇模式的推估、檢定及比較。 根據前述模式推估結果，較為普遍的模式架構一係優於模式架構二，然值得關注的是，兩種模式架構均與偏遠地區旅運者實際的選擇行為相符，顯示對偏遠地區的旅運者來說，接駁運具的選擇係比城際運具的選擇更為重要，另不同區域之車外及車內時間價值的推估結果也可支持模式推估的適合性，本研究最後則根據模式推估結果提出提升公共運輸使用率的相應策略。

In recent decades, with rapid economic development and dramatic changes in industry structures, population is highly concentrated in urban areas and the number of cars and motorcycles rapidly grows. As a result, problems of traffic congestion, energy consumption, air pollution and traffic noises jeopardize our living quality and even sustainability. To effectively tackle with these problems, it is one of the key transportation policies worldwide to encourage the usage of public transportation and reduce people dependency in private vehicles. Without knowing the traveler preferences and decisions in mode choice, it is impossible to propose effective measures to promote public transportation.

Based on this, this study attempts to identify the key variables affecting the mode choice decisions of intercity travelers so as to propose effective public transportation marketing strategies accordingly. To do so, disaggregate logit models are estimated based on a nationwide questionnaire survey. Since the procedures in mode choice decisions across areas in Taiwan (e.g. rural areas vs. urban areas) might be rather different. Two nested choice structures, Structure 1: upper level — intercity mode choice + lower level — feeder mode choice, Structure 2: upper level — feeder mode choice + lower level — intercity mode choice are respectively estimated, tested and compared. Additionally, to account for the heterogeneity of travelers living in different areas in Taiwan with rather different quality levels of public transportation system, three different models under each of structures: rural areas, urban areas, and central business districts are developed.

The estimation results show that Structure 1, the commonly adopted mode choice structure, performs better than Structure 2. However, it is worth noting that both structures are valid for the rural area model, suggesting that feeder mode choice is as important as the intercity mode choice for travelers in rural areas. Additionally, values of travel time for in-vehicle time and out-of-vehicle time in various areas are investigated to support the suitability of the estimated models. At last, corresponding public transportation promotion strategies are then proposed accordingly.