Development and Application of Controller for Transition Flight of Tail-Sitter UAV

There is renewed interest in tail-sitter airplanes on account of their vertical takeoff and landing capability as well as their efficient horizontal flight capabilities. The transition from a vertical near-hover mode to a horizontal cruise mode is a critical component of the tail-sitter flight profile. In practice, this transition is often achieved by a stall-and-tumble maneuver, which is somewhat risky and therefore not desirable, so alternative maneuvering strategies along controlled trajectories are sought. Accordingly, this paper presents the synthesis and application of a transition controller to a tail-sitter UAV for the first time. For practical reasons, linear controllers are designed using the PID technique and linked by gain scheduling. The limits of the PID controller are complemented by a so-called $\mathcal{L}_{1}$ adaptive controller that considers the coupling effect, reduces the effort for appropriate gain selection, and improves the tracking performance at different points during operation. Each transition trajectory is controlled by the flight velocity and path angle using dynamic inversion. The transition control law is tested on a tail-sitter UAV, an 18-kg vehicle that has a 2-m wingspan with an aspect ratio of 4.71 and is powered by a 100-cm³ gasoline engine driving an aft-mounted ducted fan. This paper describes not only the synthesis and the onboard implementation of the control law but also the flight testing of the fixed-wing UAV in hover, transition, and cruise modes.