Integrated predicting model for daily passenger volume of rail transit station based on neural network and Markov chain

Passenger volume prediction plays an important role in rail transit in order to support traffic planning, construction, operation, and departure interval calculation. Firstly, the paper analyzes the influencing factors of passenger volume, such as the land use. Then, by combining the characteristics of a neural network and a Markov chain, we present a predict model. The land use, information entropy, maturity of the community and the number of bus stations of the rail transit line 6, 7, 13 in Beijing were selected as the input data, and the daily passenger volume of the rail transit station as the output. The model was calibrated and verified with the data of Line 1 and 2. The result shows that neural network model can apply in long-term passenger volume prediction well and predict precision of the model are relatively high.