A probabilistic model for rail transit passenger flow distribution using AFC data

• Published in: 2022 14th International Conference on Measuring Technology and Mechatronics Automation (ICMTMA) pp. 718 - 724
• Main Authors: He, Tiejun, Long, Haidan, Yan, Zhongshu
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.01.2022
• Subjects: AFC Data; Automation; Bayesian Theory; Data models; Legged locomotion; Mechatronics; Passenger Flow Assignment; Probabilistic logic; Rails; Schedules; Urban Rail Transit
• ISSN: 2157-1481
• DOI: 10.1109/ICMTMA54903.2022.00148

It is of great significance to accurately analyze the change law of rail transit passenger flow and master the distribution status of internal passenger flow for improving the operation efficiency and service level of rail transit. The AFC (automatic fare collection) system of urban rail transit records the travel information of passengers, and massive AFC (automatic fare collection) data provides important data support for the analysis of micro passenger behavior characteristics and macro passenger flow distribution. In view of this, this paper takes micro individual passengers as the research object, and pick up passenger's ride scheme using AFC (automatic fare collection) data and train schedule data. Based on the searching results, the paper classifies the AFC (automatic fare collection) data into various trip types, and establishes the outbound walking time distribution, the inbound walking time distribution and the transfer walking time distribution respectively. Following this, the train matching models based on Bayesian theory are built to allocate the passenger flow. Finally, an example of Nanjing rail transit AFC (automatic fare collection) data is analyzed. The results show that the proposed models have good applicability and feasibility.