Research on ARIMA Model for Short-Term Traffic Flow Prediction based on Time Series

Short-term traffic flow prediction is of great significance for urban traffic management and planning. To efficiently utilize traffic flow data for prediction, it is crucial to establish a reasonable short-term traffic flow prediction model. In this study, the ARIMA method is adopted for traffic flow trend prediction. Firstly, time series data obtained from electronic police are processed to determine the stationarity of the sequence. Then, the maximum likelihood estimation method is used to estimate the model parameters, and the residuals are subjected to a white noise test. Static prediction method is used for real-time dynamic traffic flow prediction, and finally the model prediction results are evaluated. The experimental results show that the ARIMA model (4,1,5) has higher accuracy and reliability in predicting short-term traffic flow, thus it has important application value in actual traffic control and planning