Short-Term Traffic Flow Prediction Methods and the Correlation Analysis of Vehicle Speed and Traffic Flow

Abstract

A technique based adaptive neuro-fuzzy inference system (ANFIS) has been applied into short-term traffic flow prediction. Considering random factors, a new forecast model is presented based on the dynamic traffic flow data. Firstly the actual observations were operated in traffic flow and vehicle speed on Yanan east road tunnel, then the correlation analyses were calculated for traffic flow and vehicle speed in different driveways. The results shown, the correlation of traffic flow and vehicle speed was unconspicuous, but the traffic flows of the same direction in different driveways were correlated significantly. Finally according to the statistic results, some suggestions were presented.