Position control of a tail-sitter UAV using successive linearization based model predictive control

Abstract

A successive linearization based model predictive control (SLMPC) method is proposed to control a vertical take-off and landing (VTOL) tail-sitter unmanned aerial vehicle (UAV) in hovering flight. The dynamic model of the vehicle is derived, including a low-fidelity aerodynamic model and a propulsion system model. The position controller is developed by a state–space prediction model augmented with estimated disturbance and feedback integration terms. The time-varying weight in the objective function is included and the velocity of vehicle is considered as reference to improve the performance. The system is first tested in a software-in-loop environment followed by the real-time indoor flight tests. The results demonstrate the vehicle can precisely follow a trajectory and stably hold position under unsteady wind disturbance