Experimental validation of a uniformly semi-globally exponentially stable non-linear observer for GNSS- and camera-aided inertial navigation for fixed-wing UAVs

This paper provides experimental validation of a uniformly semi-globally exponentially stable (USGES) non-linear observer for estimation of attitude, gyro bias, position, velocity and acceleration of a fixed-wing Unmanned Aerial Vehicle (UAV). The available sensors are an Inertial Measurement Unit (IMU), a Global Positioning System (GPS) receiver, a video camera, and an inclinometer. The UAV is flown with the sensor payload and all data is stored locally on a hard drive, which is recovered at the end of the flight. The nonlinear observer is then tested offline with the recorded sensor data. An optical flow algorithm is used to calculate the UAV velocity based on the camera images, which is used as a reference vector of the body-fixed velocity in the attitude observer. The results are compared with an Extended Kalman Filter (EKF) and illustrate that the estimates of the unmeasured states converge accurately to the correct values, and that the estimates of the measured states have less noise than the measurements.