Attitude estimation for cooperating UAVs based on tight integration of GNSS and vision measurements

This paper presents a cooperative navigation approach for Unmanned Aerial Vehicles (UAVs) that allows robust and accurate attitude determination for a chief vehicle flying in formation with other deputy UAVs. The proposed method is based on a tightly coupled Extended Kalman Filter (EKF) that exploits the spatial diversity of measurements coming from Global Navigation Satellite Systems (GNSS) receivers and a vision system, which are integrated with inertial and magnetic sensor data. The focus is set on outdoor environments and the innovative idea is to extend attitude estimation approaches based on multiple GNSS antennas, to a multi-vehicle system where differential-GNSS and vision-based UAV-to-UAV tracking are exploited to build a virtual additional navigation sensor. Concept and processing architecture are described with emphasis on the EKF measurement update phase which is applicable for any number of cooperating deputies, and for different GNSS processing architectures. Performance of the proposed method is assessed through experimental tests involving two multi-rotors and two fixed ground antennas, one of which is used as Ground Control Point for pointing accuracy analysis. Results show the potential of the developed approach in terms of accuracy and capability to provide drift-free estimates, in real time or in post processing scenarios.