A Data-Fusion based Method for Identifying Individual Passenger Travel Dependence on Public Transportation

Priority development of public transport is an important way to implement the sustainable development of urban transport, how to scientifically identify different travelers' dependence on public transportation is conducive to explore the travelers' usage behavior of public transport, provide more accurate public transport services. Based on the dynamic and static data of public transport and individual travel survey data, this study uses the relevancy and matching technology to generate the individual travel chain information based on fused data, then selects 8 identification indicators from the dimensions of individual traveling habits behavior and individual attributes to describe the individual travel dependence on public transport. The two-step clustering algorithm which can deal with mixed variables is taken as an identification model of individual travel dependence on public transport. Finally, the identification model is applied to the actual research in Beijing, and the investigated population is clustered for four categories from the perspective of public transport travel dependence. Then the individual category of respondents is identified based on incomplete identification indicators, and the effects of assistant indicators on identification results are quantitatively evaluated with average hit ratio (AHR) and average coverage ratio (ACR). The results indicate that occupation, vehicle ownership, and income can be taken as assistant factors when the information of assistant indicators is incomplete and large scale of traveler data need to be collected and processed. The identification method of individual travel dependence on public transport can provide a meaningful reference for optimizing public transport system and improving public transport sharing rates.