Neuro-adaptive Trajectory Control of Unmanned Aerial Vehicles

A neuro-adaptive approach for autonomous flight controller design for aerial robots is proposed. Three intelligent modules are implemented to control respectively the altitude, airspeed and roll angle of the airplane, through which the altitude and the latitude-longitude of the unmanned aerial vehicle are controlled. Each intelligent module consists of a conventional feedback controller and a neural network feedback controller. The former is provided both to guarantee global asymptotic stability in compact space and as an inverse reference model of the response of the controlled system. The proposed approach makes direct use of the variable structure systems theory. A variable structure systems-based on-line learning algorithm is developed and applied to the neural network controller. Results from simulated trajectory control of the Aerosonde unmanned aerial vehicle by using the proposed neuro-adaptive control scheme are presented.