DC-FLMPC: An Online Ultimately Bounded 3-D Path-Following Controller of Fixed-Wing UAVs Under Wind Disturbance

This article studies the path-following controller for low-speed fixed-wing unmanned aerial vehicles (UAVs) with environmental wind and input limitations. Anovel disturbance-compensated fast Lyapunov-based model predictive control (DC-FLMPC) framework is proposed for fixed-wing UAVs, representing the first Lyapunov-based model predictive control (LMPC) application specifically designed for fixed-wing UAV path-following with wind disturbance compensation. First, combined with the estimation of the environmental wind by a nonlinear disturbance observer (NDO), an LMPC is employed to explicitly consider the practical constraints, such as input and practical asymptotic stability constraints. Second, a refined continuation/generalized minimal residual (rC/GMRES) algorithm is developed with smooth maximum functions to handle inequality constraints, enabling real-time solution of stability-constrained optimization problems. Third, an adaptive horizon adjustment strategy dynamically adjusts the prediction horizon length based on the rate of change in tracking error, saving computational resources while maintaining stability. More importantly, the backstepping control (BC) technique constructs NDO-based stability constraints ensuring practical asymptotic stability under wind disturbance with rigorous theoretical guarantees. Furthermore, the recursive feasibility and closed-loop stability of the DC-FLMPC algorithm are theoretically analyzed. Finally, numerical simulations are given to demonstrate the effectiveness of the proposed algorithm.