Since the origin-destination and transfer information calculation by automatic fare collection data under large-scale subway network always take a long time, this paper proposes a method to extract subway trip information according to cell phone data triggered principle and adjacency matrix of subway stations. First, cell of origin is used to identify entering station after pre-processing, time difference of arrival is used to identify exit position according to the last recognized transfer station and a group of nearest stations. Then the adjacency matrix about stations and lines is established, the two types of transfer stations are distinguished based on triggered principle between same-floor and different-floor transfer. Finally, the extraction accuracy of trip information is verified by AFC data and cell phone data in Chongqing. The identification rate of the entering station is 91.18%; the same-floor transfer station is 88.36%; the different-floor transfer station is 87.85%; the exiting station is 90.21%.

CICTP