Integrated Multi-Sensor State Estimation Using GNSS and Redundant IMUs for UAVs Beyond Visual Line of Sight

This paper proposes a filter architecture which incorporates accelerometer and gyroscope measurements from redundant IMUs into the state estimation of UAVs. A loosely coupled, federated extended Kalman filter combining Galileo, GPS, barometer, magnetometer and measurements from two IMUs is presented. A navigation sensor configuration for a UAV of class 2 will be presented. The proposed filter algorithm and the sensor configuration are tested and validated using an ultralight aircraft equipped with highly precise reference equipment. This paper concludes with an analysis of the achievable accuracy and the improvements made by using both IMUs for the state estimation.