Plenty of drones are capable of withstanding mild winds. Things get a bit more complicated in extreme weather though. Getting autonomous drones to learn how to move on a moving platform in turbulent wind conditions comes with unique challenges. AerospaceControlsLab has shared a video that shows how drones can be enhanced with a vision system to tackle the problem better.
As the researchers explain:
a fully autonomous vision-based system that addresses these limitations by tightly coupling the localization, planning, and control, thereby enabling fast and accurate landing on a moving platform. The platform’s position, orientation, and velocity are estimated by an extended Kalman filter using simulated GPS measurements when the quadrotor-platform distance is large, and by a visual fiducial system when the platform is nearby. The landing trajectory is computed online using receding horizon control and is followed by a boundary layer sliding controller that provides tracking performance guarantees
You can find the paper here.