Vehicle future trajectory prediction with a DGPS/INS-based positioning system

This paper explores the feasibility of vehicle future trajectory prediction when vehicles are equipped with a DGPS-based positioning system. A major advantage of the DGPS-based positioning system is that it provides vehicle state information (positions/velocities) in the Earth inertial coordinates. This advantage facilitates the prediction of vehicle future trajectories, and the usage of the positions and trajectories of nearby vehicles for the construction of the local driving environment. This paper examines possible methodologies in the trajectory prediction, and analyzes the error statistics of the prediction by treating the model-based prediction as an equivalent open-loop Kalman filter. The trajectory prediction is then applied to the development of a cooperative collision warning system; experimental results show that the trajectory prediction achieves sufficient accuracy for the application.