Taxis are the most important components for public transit systems. The model of taxis is more flexible and convenient for passengers compared to other transit model like buses or railway. However, the dynamic of taxi locations will have influence on transit efficiency. Therefore, the hotspot selecting patterns when unoccupied with regards to different income level drivers were investigated in this study. Two month GPS data of 7200 taxis in Wuhan was used as the data samples. The Wuhan City was firstly Mapped Meshing and divided into 4623 grids. After preprocessing the taxi GPS data and taxi drivers' income level classification (top, ordinary and bottom), the pickup points were filtered and matched with all these grids. Heat and grid probability for passenger demand hotspot were proposed and analyzed in this study. The results showed that the correlation value between top drivers and heat are not always higher than ordinary taxi drivers but the correlation value between grid probability and top taxi drivers are always high at all time slots. The finding of this study reveals that the high income taxis drivers have the high ability for cruising to the closed hotspot having high grid probability compared with middle or low income drivers. It therefore recommended that, if such experience could be represented and learned by other taxi drivers, the efficiency of taxi transit systems will be improved. Furthermore, if the information of real time heat and grid probability of hotspot could be broadcasted to taxi drivers, it will be beneficial for taxi transit systems as well.
Research on taxi drivers' passenger hotspot selecting patterns based on GPS data: A case study in Wuhan
2017-08-01
923664 byte
Conference paper
Electronic Resource
English