Robust observer-based backstepping controller for a quadrotor UAV

Abstract

A novel robust tracking controller for a quadrotor UAV is designed that is capable to compensate the unknown unmatched uncertainties in the system model. A backstepping approach is employed in which such perturbations are estimated on-line by a nonlinear disturbance observer (NDO). The NDO is employed to estimate the effect of wind perturbations and unknown aerodynamical forces. Global asymptotic tracking of the desired position and yaw angle of the quadrotor is achieved in closed loop where the controller gains are selected to satify conditions that follow from stability analysis of the system. The system performance exhibits much better robustness than the existing backstepping control methods which are not equipped with nonlinear disturbance estimators.