Quadrotor Control

Quadrotor control concerns keeping these agile aerial robots stable while tracking trajectories, avoiding obstacles, and handling disturbances, changing payloads, or physical interaction. Common approaches include geometric and PID controllers, incremental nonlinear dynamic inversion, and model predictive control, often paired with control barrier functions for safety; learning methods such as reinforcement learning and learned dynamics are increasingly used for agile flight and adaptation. Applications range from racing and cluttered navigation to inspection, landing, formation flight, and transporting or manipulating payloads. A central challenge is achieving reliable sim-to-real transfer and safe performance under uncertain dynamics while meeting the tight computation and sensing limits of small onboard platforms.