Conference article

Passenger Flow Prediction of High Speed Railway Based on LSTM Deep Neural Network

Jie Li
School of Transportation and Logistics, Southwest Jiaotong University, High-tech Zone West Campus, Chengdu, China / Civil, Buildings and Environmental Engineering Department, SAPIENZA Università di Roma, Roma, Italy

Ping Huang
School of Transportation and Logistics, Southwest Jiaotong University, High-tech Zone West Campus, Chengdu, China

Yuxiang Yang
School of Transportation and Logistics, Southwest Jiaotong University, High-tech Zone West Campus, Chengdu, China

Qiyuan Peng
School of Transportation and Logistics, Southwest Jiaotong University, High-tech Zone West Campus, Chengdu, China

Download article

Published in: RailNorrköping 2019. 8th International Conference on Railway Operations Modelling and Analysis (ICROMA), Norrköping, Sweden, June 17th – 20th, 2019

Linköping Electronic Conference Proceedings 69:47, p. 723-739

Show more +

Published: 2019-09-13

ISBN: 978-91-7929-992-7

ISSN: 1650-3686 (print), 1650-3740 (online)

Abstract

The paper presents the characteristics of the departing passenger flow in different stations based on the real-record passenger flow data of Wuhan-Guangzhou high speed railway, from January, 2010 to December, 2015. The passenger dataset is framed for the long short-term memory (LSTM) model, considering the expectation input format of LSTM layers and the characteristics of the data. The Keras model in Python is used to fit LSTM model with tuning and regulating all the parameters necessary in the model. Then the fitted LSTM model is applied to forecast the short-term departing passenger flow of Wuhan-Guangzhou high speed railway. The influence of important parameters in the LSTM model on the prediction accuracy is analysed, and the comparison with other representative passenger flow forecast models is conducted. The results show that the LSTM model can get the valid information in a long passenger flow time series and achieve a better performance than other models. The passenger flow prediction errors valued by MAPE are 7.36%, 7.33%, 8.03%, respectively for Chenzhou station, Hengyang station and Shaoguan station. The parameters in the LSTM model such as the number of hidden units, the batch size and the input historical data length have a great influence on the prediction accuracy.

Keywords

High speed railway, Passenger flow prediction, Long short-term memory model, Deep learning, Time series

References

No references available

Citations in Crossref