Adaptive neural network control for aircraft landing under actuator failures

Neural network (NN) based adaptive control is investigated for aircraft landing in the present of actuator failures in this work. Radial basis function neural network (RBFNN) in combination with backstepping technique is utilized to develop the virtual control for the overactuated aircraft with dynamic uncertainties and unknown disturbances. By utilizing Lyapunov synthesis, the closed-loop control signals are proved to be semi-globally uniformly ultimately bounded, and the tracking error converges to a small neighborhood of origin. Control allocation (or reallocation once actuator failures have been detected and isolated) is designed to distribute the virtual control among the redundant actuators with amplitude and rate constraints. Simulation results are presented to show the effectiveness of the proposed control.