Research on robust GNSS vehicle three-dimensional tracking method for urban elevated road networks

The Kalman filtering (KF) has been implemented as the primary scheme for many land vehicle navigation and positioning applications. However, it has been reported that the KF-based techniques have limitations that it assumes the noise is Gaussian white and the system model must be known exactly. Due to the complicated vehicle tracking environment in urban area (signal disappear, attenuation or reflection) and diverse vehicle motion (uniform or accelerated), The VNS inevitably exits stochastic uncertainties whose statistical property can not be priori known. This makes great difficulties in tracking vehicle robustly. In this paper, robust GNSS vehicle three-dimensional tracking method for urban elevated road networks is investigated. By exploring the geometry of the vehicle tracking problem, the three-dimensional vehicle tracking problem is formulated to one-dimensional target trajectory tracking problem. Accounting for modeling uncertainties and unpredictable disturbances problem, via robust H^∞ filtering algorithm with Stochastic Uncertainties that we have developed in another paper, a three-dimensional vehicle tracking algorithm for urban elevated road networks is proposed. The experiment results confirm the effectiveness of the proposed method by comparing with the Kalman filter tracking method using the measured GNSS data.