模化旅運者運具選擇行為之動態性與異質性

Abstract

近年來全球暖化已為各國關注議題，根據官方統計資料顯示，台灣運輸部門二氧化碳排放量僅次於工業部門。無庸置疑的，綠色運輸系統導向政策不僅為提升公共運輸使用率且同時減少旅運者對私人運具依賴的關鍵政策，同時也能達到環境保護的作用。然而，依據交通部統計調查結果，目前臺灣地區僅有四個縣市公共運輸使用率大於百分之十，對於改善提升公共運具使用率還有很大的進步空間。無論如何，在未知旅運者其運具選擇行為的情況下，無法針對公共運具確切改善和提出有效策略，因此許多研究已經發展運具選擇模式以衡量旅運者選擇行為。

在以往運具選擇研究中指出，透過觀察同一位受訪者其選擇行為，能預測其慣性偏好行為。基此，本研究建構靜態和動態運具選擇模式，以評估各種不同運輸政策來促進公共運輸使用率。此外，本研究進一步依照旅運者所居住鄉鎮市區之人口密度將其做區分，分別為偏遠區、都市區和都會區並建構運具選擇模式，以達因地制宜效果。為探討旅運者之異質性，研究利用潛在類別模式將旅運者做適當市場區隔，探討區隔間旅運者的特徵，並根據潛在類別模式推估結果，針對不同區隔提出管理策略。本研究針對全國機動車輛持有家戶，進行大規模運具選擇行為追蹤問卷調查，其旅次特性為針對區內旅次(單程旅次距離大於500公尺、小於20公里)之運具選擇行為。

結果顯示在這三個地區當中，車內時間(變化量)及旅行成本(變化量)此兩個變數，為顯著影響旅運者移轉所選運具之因素。最後綜合模式推估結果，進一步計算彈性與市占率，研擬適宜的公共運輸政策，以提升公共運輸使用率。

Global warming has become a major concern worldwide nowadays. According to official statistics, the amount of carbon dioxide emitted by the transport sector is ranked second next to the industrial sector in Taiwan. Towards a greener transport system, undoubtedly, to increase public transport usage rate and reduce travelers’ dependency on private vehicles is one of the key policies for achieving environmental protection and a better living environment. However, according to the survey of Ministry of Transportation and Communications (MOTC), only four cities out of a total of 22 cities/counties in Taiwan have public transportation usage rates exceeding 10%, suggesting there is still a large room for the improvement and promotion of public transportation. However, without knowing the mode choice behaviors of travelers, it is impossible to prioritize improvement and promotion strategies of public transportation, which explaining that numerous studies have been conducted to develop mode choice models.

Previous studies have indicated that observing repeated choice behavior can identify the effect of inertia. To this end, this study develops static and dynamic mode choice models to evaluate the effectiveness of various transport strategies in promoting the use of public transport. Besides, the study develops separate mode choice models for travelers living in the districts/townships with different levels of population density, including rural, urban, and metropolitan areas. To acknowledge the heterogeneity of travelers, this study further uses the latent class model to identify segment-specific preferences and traveler profiles of each segment. A large-scale panel survey data were collected from the travelers in Taiwan, who own private vehicles and had engaged a regional trip (one-way trip distance longer than 500 meters and shorter than 200 kilometers).

The estimated results show that in-vehicle travel time and total travel cost have the significan1t impacts on travelers’ choice behaviors. Finally, based on model estimation results, this study evaluates the changes in mode shares with respect to a number of transport strategies and provides policy implications in improving public transport.