Robust Adaptive Backstepping Control of UAV with Lumped Uncertainties

The design of a robust adaptive backstepping flight control law for a unmanned aerial vehicle (UAV) is discussed. The lumped uncertainties of the nonlinear UAV model are concerned in this paper. RBF neural networks are used inside the backstepping control law to approximate the lumped uncertainties. Command filters are used to implement the virtual control law and avoid "explosion of complexity". The controller and its performance are evaluated on a nonlinear, six-degrees-of-freedom dynamic model of the UAV in the simulations.