A K Nearest Neighborhood-Based Wind Estimation for Rotary-Wing VTOL UAVs

摘要

Wind speed estimation for rotary-wing vertical take-off and landing (VTOL) UAVs ischallenging due to the low accuracy of airspeed sensors, which can be severely affected by the rotor’sdown-wash effect. Unlike traditional aerodynamic modeling solutions, in this paper, we present a KNearest Neighborhood learning-based method which does not require the details of the aerodynamicinformation. The proposed method includes two stages: an off-line training stage and an on-line windestimation stage. Only flight data is used for the on-line estimation stage, without direct airspeedmeasurements. We use Parrot AR.Drone as the testing quadrotor, and a commercial fan is used togenerate wind disturbance. Experimental results demonstrate the accuracy and robustness of thedeveloped wind estimation algorithms under hovering conditions.