This paper tackles the problem of path-following control for fixed-wing unmanned aerial vehicles (UAVs), while accounting for wind disturbances and hindering obstacles. We introduce a novel predictive algorithm based on a quasi-linear parameter-varying (qLPV) model representation of the 3D kinematics of the fixed-wing aircraft. This approach allows us to utilize efficient Quadratic Programming (QP) solvers to find efficient and fast solutions to the Optimal Control Problem (OCP), typically within milliseconds. Additionally, it facilitates the incorporation of appropriate constraints aligned with the aircraft dynamics and obstacle constraints after further processing. In this paper, we demonstrate how the nonlinear obstacle constraints can also be represented in a qLPV form, making it feasible to handle them within our framework. Moreover, stability conditions can be directly derived based on the qLPV representation. The algorithm's effectiveness is demonstrated on an aerobatic unmanned aircraft with a successive-loop-closure (SLC) based attitude and stabilization controller. The evaluation is conducted across two scenarios previously used in experimental flights with the same aircraft. Each scenario involves nine waypoints, obstructive obstacles, and wind disturbances. The simulations begin with the kinematic model and are subsequently extended to a high-fidelity model of the UAV, resulting in successful path-following and obstacle avoidance with relatively low computational times.
A Velocity qLMPC Algorithm for Path-Following with Obstacle Avoidance for Fixed-Wing UAVs
2024-06-04
2452521 byte
Conference paper
Electronic Resource
English
End-to-End Learning-Based Obstacle Avoidance for Fixed-Wing UAVs
Springer Verlag | 2023
|Practical Obstacle Avoidance Path Planning for Agriculture UAVs
British Library Conference Proceedings | 2019
|