Gain-Scheduled Control of a Quadcopter UAV

Abstract

In this thesis we develop a gain-scheduled control law for the quadcopter unmanned aerial vehicle (UAV). Techniques from linear control theory are introduced and used to construct adaptive proportional and proportional-integral control laws for use with both state and observer-based output feedback. The controller monitors the yaw angle of the quadcopter and updates a gain matrix as the system evolves through operating points. To demonstrate the effectiveness of the gain-scheduled controller, trajectories involving significant variation in the yaw angle are tracked by the quadcopter, including a helix and Lissajous curve. We consider physical implementation of the controller, and offer suggestions for improvement and future work.