US Patent 11428212 Wind turbine drivetrain wear detection using azimuth variation clustering

Systems and methods to monitor a wind turbine azimuth drivetrain. Azimuth variation characteristics data are accumulated from wind turbines over a period of time. Clusters of values within the azimuth variation characteristics data are identified and a respective condition of the main drivetrain is associated with different clusters of values. After the associating, a measured set of azimuth variation characteristics data is received. A cluster corresponds to values in the measured set of azimuth variation characteristics data is determined and a condition associated with that cluster is determined to be a condition associated with the subject main drivetrain. That condition is then reported.