Real-time Estimation of Urban Rail Transit Passenger Flow Status Based on Multi-source Data

Abstract

The AFC data uploaded in real time for urban rail transit is incomplete and delayed. In order to improve the dynamic management and control level of rail transit, this paper describes the real-time state estimation process of passenger flow and its key issues, and analyses the relationship between real-time uploaded AFC data and mobile phone signal data, establishing a multi-source data fusion model based on gradient descent method; On this basis, combined with the multi-source data after fusion as the basis of real-time estimation, a realtime passenger flow estimation model based on Kalman filter is established. Practice shows that the fusion of multi-source data improves the accuracy of estimating basic data, the error of the real-time passenger flow estimation model based on Kalman filter is less than 10%, and the accuracy of the estimation model is good. The needs of the urban rail transit operation management department to grasp the real-time passenger flow distribution status of the network are met.