Nonlinear model predictive control and guidance for a propeller-tilting hybrid unmanned air vehicle

Abstract

Hybrid unmanned aerial vehicles (UAVs) provide an interesting combination of the vertical takeoff and landing (VTOL) capabilities of rotary-wing (RW) and of the efficient forward flight of fixed-wing (FW) vehicles. The employed controllers must be able to handle the highly nonlinear dynamics and changing control authorities resulting from this combination, especially during the transition between the two flight modes. In this paper, a nonlinear model predictive control (MPC) structure is designed and applied to a tiltrotor convertible UAV. Full-flight envelope trajectory tracking and optimal exploitation of the aircraft’s VTOL and FW properties is achieved. A common approach is to design a set of stable controllers for various trim-points in the flight envelope and make use of Gain Scheduling (GS) or controller mixing, in order to select the appropriate control law for a given flight configuration. In this paper, a unified control approach is developed. A nonlinear MPC structure is designed and applied to a tiltrotor convertible UAV whose nonlinear model is derived and presented. Thus, full-flight envelope trajectory tracking and optimal exploitation of the aircraft’s VTOL and FW properties are achieved, without the need for controller switching or scheduling policies. The proposed multi-stage control allocation handles the changing control authorities of the actuators in a continuous manner and efficiently distributes the required control actions between propellers, tilt servos and control-surfaces. The feasibility and performance of this novel control approach are successfully evaluated in real-world flight experiments. Simulation results for comparison with state-of-the-art approaches are presented.