Short-term passenger flow prediction of urban rail transit based on ARIMA model

Abstract. This paper is about the short-term passenger flow prediction of urban rail transit based on the ARIMA model. Due to the acceleration of urbanization and the increasing density of urban rail transit, more and more people are taking urban rail transit. However, excessive short-time passenger flow can lead to traffic accidents. Therefore, it is very important to predict short-term passenger flow. This paper aims to use the ARIMA model to predict the passenger flow in a short period of time in urban rail transit. A competition data set from Alibaba about the short-time passenger flow of a metro station in Shenzhen was adopted and analyzed by the ARIMA model. The data set contains 1,428 samples, and the time granularity is 10 minutes. After research, it has been found that the ARIMA model can basically predict the short-time passenger flow of the metro with accuracy under the appropriate parameter setting. This also means that the application of the ARIMA model can be extended to other aspects of short-time passenger flow prediction, not just the metro.