Traffic Flow Analysis and Prediction Based on GPS Data of Floating Cars

Due to the limited amount of sensor infrastructure and its distribution restrictions, it is a challenge for the traffic related groups to estimate traffic conditions. The GPS log files of floating cars have fundamentally improved the quantity and quality of wide-range traffic data collection. To convert this data into useful traffic information, new traffic models and data processing algorithms must be developed. To conduct a comprehensive and accurate traffic flow analysis, this paper proposes a traffic flow analysis and prediction process based on GPS data of floating cars. For the key links involved in the process, new models and algorithms are implemented. Based on the actual data of a city, we validate the effectiveness of the design process, and provide a reliable solution for urban traffic flow analysis.