Wind-Driver: A Framework for High-Maneuverability Planning and High-Precision Control for Quadrotor Autopilots

In nature, albatrosses can skillfully leverage environmental winds to fly over long distances, thus conserving energy. This bio-inspired approach highlights the potential for quadrotors to utilize wind. When a quadrotor flies amidst wind fields and obstacles, the wind can be utilized to enhance flight safety and maneuverability. This paper presents a coupled trajectory planning and control framework named Wind-Driver for quadrotor autopilots. A refined quadrotor dynamic model is established starting from the blade element momentum theory. Based on the refined model, the integral-based wind speed observer is designed to guarantee lower latency and more accurate estimation thus improving tracking performance. The estimated wind speed and force are utilized in the planning layer. The estimated wind speed is leveraged to establish the wind-risk-aware A* algorithm in front-end path search, which guides the quadrotor to select the wind-sheltered side of obstacles. Using the external wind force, the wind utilization trajectory optimization design is pursued to relax the dynamic constraints in the back-end. The generated trajectory is transmitted to the controller for tracking. Simulations and flight experiments are carried out to verify the robustness and practicality of the proposed framework. Benchmark comparisons and analyses indicate that the proposed approach outperforms other algorithms.