Passenger Flow Forecast of Urban Rail Transit Based on BP Neural Networks

Firstly, according to the Beijing urban rail transit network characteristics and based on the data of the historical passenger flow, the passenger flow in sections is distributed and the referenced passenger flow in sections is gotten on the theoretical basis of the shortest path distribution of static unbalanced distribution model. Then through a lot of BP neural network modeling experiments, a reasonable prediction model is established to aim at Beijing urban rail transit passenger flow forecast problem. Finally, through using the BP neural network model, transfer passenger flow in sections is predicted from Fuxingmen to Fuchengmen station on Beijing urban rail transit Line 2 and reasonable passenger flow forecast results are gotten to prepare for passenger traffic scheduling system research.