S.W. Li, Y. Li, J. Yang, C.T. Cao, J.F. Yang, H.D. Zhou
In the one-entrance-ticket (single-pass) public transportation system, a passenger only swipes the card when getting on the bus, but does not swipe the card when getting off. Consequently, information on the time and the station where the passenger gets off cannot be obtained, and passenger travel OD (Origin-Destination) cannot be determined. This situation leads to the need for a method of calculating the getting-off station through the getting-on station and time. This research presents methods of calculating passenger getting-off station with regular traveling (going to work and getting back home on weekdays) and closed bus trip chain (the passenger takes the bus as the basic vehicle for all the trips during the day). Moreover, the study merges, deals with, and analyzes bus IC card data, bus GPS data, and bus station and route, and other GIS data. The study obtains suspected getting-off station information by conducting spatio-temporal simulation of passenger traveling characteristics of bus IC cards and buffer analysis of getting-on stations. Data for analysis include historical matching rate of suspected getting-off stations, mean distance between nearby key POIs (point for interest), number of transfer times, getting-on station distance (road network and spatial distance), and other factors. The passenger getting-off stations are calculated and reliability of the calculation method is verified. The focus of the study is the Guangzhou Sports Center station. Results indicate that in studying passenger getting-off stations, the algorithm used accurately matched the passenger getting-off stations recorded in the questionnaire survey.
Keywords: passenger travel OD; GPS data; IC card data; data fusion; GIS spatio-temporal simulation; distance factor