Adaptive nonlinear control for tiltrotor aircraft

The combination of feedback linearization and adaptive 'neural' networks provides a powerful controller architecture. This paper presents some of the nonlinear, and adaptive flight control research being conducted in this area in the School of Aerospace Engineering at the Georgia Tech. It includes adapting-while-controlling neural networks, which are guaranteed to remain bounded. The application is unique in the sense that it allows for uncertainty and nonlinearities in control, as well as in states. A description of the controller architecture and associated stability analysis is given. This is followed by its application to a tiltrotor aircraft, using the nonlinear Generic Tiltrotor Simulation code developed NASA Ames Research Center.