Passenger Trajectory Reduction in Urban Rail Transit Station Based on Probing Data

With the increasing requirements for the analysis of passenger flow and the increasing coverage of Wi-Fi signals in rail transit stations, the high precision advantage through Wi-Fi probing data becomes increasingly prominent in obtaining passenger trajectory. The research on trajectory mining is mature in outdoor environment. However, for indoor environment there are massive noise in Wi-Fi probing data, which significantly interferes the preciseness of the trajectory reduction and brings great difficulty to existing trajectory reduction methods. To this issue, the paper proposes a novel passenger trajectory reduction framework for urban rail transit system, which is composed of trip trajectory division, trajectory noise data cleaning, and semantic trajectory extraction. In addition, the system considers the spatial topology characteristics of the rail transit station. Realistic trajectory Wi-Fi data from Hanzhong Road Station of Shanghai Metro is utilized to support the experiments. The results demonstrate that the proposed method can mine the space-time trajectory from the original noise trajectory data efficiently and accurately, and successfully provide support for passenger flow analysis and station streamline optimization.