Real-Time Traffic State Estimation and Short-Term Prediction Using Probe Vehicle Data: A Kalman Filter Approach

This paper presents a comprehensive Kalman filter methodology for real-time traffic state estimation and short-term prediction using probe vehicle data. The approach addresses the challenges of sparse and irregular probe vehicle data in traffic state estimation applications. The methodology demonstrates significant improvements in prediction accuracy and provides reliable real-time traffic information for intelligent transportation systems. The system is validated using real-world data and shows promising results for deployment in traffic management applications.

Recommended citation: A. K. Shafik, and H. A. Rakha, "Real-Time Traffic State Estimation and Short-Term Prediction Using Probe Vehicle Data: A Kalman Filter Approach," in Sensors (In Press).
Download Paper | Download Slides