Passenger Flow Forecast of Metro Station Based on the ARIMA Model

By the end of 2014, 83 metro lines with a length of over 2500 km in total had been constructed in 22 metropolitan cities in mainland China. A series of worth exploring and pondering problem arises in the construction process, and the passenger flow prediction analysis of metro station is one of them. This paper built an ARIMA model which is a kind of short-time traffic forecasting model with high precision. The detailed data of historical passenger flow in section in a typical station are fitted in this paper. On the basis of this, the passenger flow in the next day is forecasted and analyzed. The fitting is with the help of statistical software called SPSS. Finally, the model of ARIMA (3, 0, 2) is built up. The results showed that the ARIMA model prediction has certain accuracy. It can solve the problem of modeling about non-stationary time series prediction.