Prediction of Passenger Flow Based on CNN-LSTM Hybrid Model

Abstract

Passenger flow forecasting is an inevitable need of modern railway operation, and also a hot and difficult research topic in the field of railway transportation for a long time. Accurate passenger flow forecasting can effectively allocate transport resources, meet passenger travel needs, and improve the operational and social benefits of railway enterprises. Inspired by the development of Deep Learning in the field of natural language processing and image recognition, this paper proposes a hybrid model of CNN-LSTM, which is applied to forecast the dispatched passenger number in the section from Beijing West to Zhengzhou East. In this paper, the key factors (variables) affecting the model, such as short-term factors, long-term influencing factors, are considered comprehensively. It is hoped that the correlation of sample data in time series dimension can be maximized. Finally, CNN model is introduced to further extract the characteristics of sample data, so as to optimize and improve the prediction accuracy of the model. The results show that the error of CNN - LSTM hybrid model is 4.92% and 6.58% under MAE and RMSE standards, which is obviously better than ARIMA model and single LSTM model.