An adaptive robust nonlinear observer for velocity and disturbance estimation for the translational motion of a quadrotor aircraft

An adaptive robust observer for UAVs position estimation problem is proposed in this paper. The observer design was obtained from a change of coordinates, where a nonlinear system was transformed to a canonical observable model. A practical stability was proved using the Lyapunov approach, and it is proved that the states converge inside a ball that depends on the high gain radius r, which is given in the proof. Simulation scenarios were carried out to validate the performance of the closed-loop system. These validation experiments were under ideal classical conditions without disturbance and then disturbance was added in the states to emulate a real system. Results showed the adaptive disturbance rejection of the observer. In addition, the observer was also validated online by measuring real data from outdoors flight test to estimate the velocity and disturbances of translational movement.