Performance evaluation of lane level positioning based vehicle collision avoidance for cooperative vehicle infrastructure system

Accurate and real-time vehicle positioning, especially at the lane level, is essential for many of the applications of cooperative vehicle infrastructure system (CVIS), which is a prospective trend of future intelligent transportation systems. Integration of on-board sensors, a digital road map database and support from V2V communication is critical to maximize the benefits to achieve lane-level precision for specific applications, in particular collision avoidance for traffic safety assurance. In this paper, design of the lane-level positioning based vehicle collision avoidance system (VCAS) is presented, the issue of multi-sensor fusion filtering is discussed, and then the vehicle collision avoidance strategy for unsignalized intersections is detailed and analyzed. Practical road map measurements are employed to build scenarios with simulated vehicle trajectories, by which simulations are carried out to evaluate impacts of positioning and communication qualities for the proposed VCAS scheme and solutions. The simulation results demonstrate the capabilities of vehicle safety assurance and efficiency improvement for specific cooperative vehicle infrastructure system applications.