Passenger Flow Forecast Algorithm for Urban Rail Transit

Abstract

To exactly forecast the urban rail transit passenger flow, a multi-level model combining neural network and Kalman filter was proposed. Firstly, ELAN neural network model was introduced to implement a preliminary forecast of the passenger flow. Then the Kalman filter was used to correct the preliminary forecast results, so as to further improve the accuracy. Finally, in order to validate the proposed model, the passenger flow in Shanghai subway transport hub was observed and simulated. Experimental results showed that the proposed multi-level model reduced error by about 0.8% and had better actual effect compared with any single algorithm.

DOI : http://dx.doi.org/10.11591/telkomnika.v12i2.3810