This project proposes to enhance a decentralized traffic signal controller based on a Nash Bargaining game-theoretic framework, integrating a Kalman filtering (KF) algorithm for real-time traffic state estimation. By combining KF with the traffic signal controller, the project aims to optimize signal phasing sequences at intersections based on turning movements and traffic density, thereby reducing queue lengths and delays. The approach involves traffic data collection through loop detectors and probe vehicle data, and it will address saturation flow rates for shared lanes. This integration intends to achieve efficient system performance across varying probe vehicle market penetration levels, ultimately providing a robust solution for improved traffic flow at intersections.
Universities Involved
Virginia Tech
Principal Investigators
Hesham A. Rakha
Amr Shafik
Expected Research Outcomes & Impacts
The study expects to demonstrate reduced congestion at signalized intersections by optimizing traffic flow, vehicle delay, and queue length. The results will quantify the improved performance of the integrated KF and game-theoretic model across various levels of data availability, offering an adaptable approach for real-world applications in eco-driving and traffic management.
Subject Areas
Traffic Management, Signal Phase and Timing