Invariant Extended Kalman Filter Applied to Tracking for Air Traffic Control
Marion Pilté, Sami Jouaber, Bonnabel Silvère, Frédéric Barbaresco, Marc Fragu, Honor Nicolas · 2019
In this paper, a recent target model expressed in intrinsic coordinates and based on the Frenet-Serret frame is reviewed. The associated variant of the EKF suited to track the corresponding state, and called the Invariant Extended Kalman Filter (IEKF) is also reviewed. Our main contribution is to apply those recent tools to Air Traffic Control (ATC) situations. For ATC application, a good precision on the velocity vector is critical, especially its direction (i.e., the heading of the plane), to display a coherent view of the situation in the sky to the operator. This filter is implemented in a realistic ATC tracking chain to test the algorithm under realistic conditions, when inserted in the complete tracking pipeline. The new filter proposed is then compared to an Extended Kalman Filter (EKF) with a constant velocity target model, which is fairly simple but widely used in practice for ATC. We show the results are improved, especially during turns .