Short-term Subway Passenger Flow Prediction based on LSTM and ARIMA Model

With the acceleration of urbanization, urban traffic problems have become increasingly severe. To alleviate urban traffic congestion, the development of public transportation has been greatly emphasized. As a convenient and rapid means of urban travel, the metro has become the preferred choice for citizens, leading to significant passenger flow pressure. Therefore, it is crucial to forecast passenger flow in advance. This paper utilizes passenger flow data from Chengdu Metro Line 1 from May to September 2019 to conduct a study. LSTM and ARIMA models were constructed to simulate the actual passenger flow data. The prediction performance of both models shows that the trends simulated by these models fit the original data well, verifying their applicability in passenger flow forecasting. Furthermore, it was found that the LSTM model outperformed the ARIMA model among the constructed models in this paper. Accurate metro passenger flow forecasting can help relevant departments manage the metro system more effectively, thereby improving passenger travel efficiency and safety.