AI techniques applied to diagnosis of vibrations failures in wind turbines

Abstract

Supervision and fault diagnosis in wind turbines using automatic learning techniques allow early detection of the degeneration of the components, as well as the detection and diagnosis of sudden failures. This contribution evaluates different machine learning methodologies to predict, detect and diagnose electrical and mechanical failures of wind turbines. An integrated monitoring and diagnostic system is proposed using automatic learning algorithms adapted to the different components and faults of the wind turbine.