Data–driven Modelling of a Wind Turbine Benchmark for Fault Diagnosis Application

Abstract: In order to improve reliability and safety of wind turbines, it is important to detect and isolate faults as fast as possible, and handle them in an optimal way. This work describes a data–driven modelling approach oriented to the design of a diagnosis scheme, used to detect faults, and isolate them as early as possible, in order to avoid possible catastrophic consequences. A hybrid modelling approach is used here since the model under investigation is nonlinear, whilst the wind speed measurement is uncertain since it is influenced by the turbulence around the rotor plane. The modelling method relies on piecewise affine prototypes, which are identified from the noisy measurements acquired from the simulated wind turbine. The fault detection and isolation strategy is thus designed based on these hybrid models. The wind turbine simulator is finally used to validate the achieved performances of the suggested fault diagnosis scheme.