Objective  Subways are composed of three core components: passengers, stations, and trains. By constructing the dynamic relationship between these three components based on OD (origin - destination) and train timetable information, a backtracking study of network passenger flow is conducted.

Method Current research methods for network passenger flow backtracking are analyzed. A network passenger flow backtracking model is constructed based on passenger OD and train timetable information. This model is validated through experiments using actual data from Hangzhou Metro.

Result & Conclusion  Based on the real data from Hangzhou Metro, calculation results of the constructed network passenger flow backtracking model show that the average platform waiting time for passengers is 2.88 min, the average error in passenger exit time is 3.27 min, and the average passenger travel time is 26.50 min, indicating an accuracy rate of over 88% for this model.