In this paper, a destination choice model is developed for taxi passengers based on taxi GPS data from Shanghai, China. Taxi GPS data belong to passively collected big data that can avoid possible biases in traditional travel surveys limited by the sampling process, and potential discrepancies between respondents’ actual behaviors and their responses. As a discrete choice model, a destination choice model can involve policy-sensitive variables in a flexible way, so as to predict and evaluate policy impacts. The developed model incorporates a variety of explanatory variables, including travel impedance variables (travel time and monetary cost), location indicator variables (whether an airport or passenger railway station is in the traffic analysis zone, i.e. TAZ) and attraction variables in trip destination ends (population and employment). Finally, the factors influencing taxi passengers’ destination choice behaviors are analyzed based on the model estimation results.
Development of Destination Choice Model for Taxi Passengers in Shanghai, China
17th COTA International Conference of Transportation Professionals ; 2017 ; Shanghai, China
CICTP 2017 ; 3989-3997
18.01.2018
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Taxi Passengers Demand Prediction Using Deep Learning
TIBKAT | 2017
|Method for deducing waiting time of cruising taxi passengers
Europäisches Patentamt | 2023
|British Library Conference Proceedings | 2007
|