Fault Diagnosis of Offshore Wind Turbines via Identified Fuzzy Residual Generators

In order to improve the safety, the reliability, the efficiency, and the sustainability of offshore wind turbine installations, thus avoiding expensive unplanned maintenance, the diagnosis of faults in their earlier occurrence is fundamental. Therefore, the main contribution of this work consists of the development of a model–based fault diagnosis scheme applied to wind turbine nonlinear models. In particular, a data–driven strategy relying on fuzzy models is exploited to build the residual generation scheme. Fuzzy theory is exploited here since it allows to approximate easily unknown nonlinear models and manage uncertain data. Moreover, these fuzzy models, which are directly identified from the wind turbine measurements, lead to the direct design of the fault diagnosis module. It is worth noting that, in general, the nonlinearity of wind turbine systems would generate complex analytic models. This aspect of the work, followed by the simpler solution relying on IF–THEN fuzzy rules, represents the key point when on–line implementations are considered for viable applications of the proposed methodology. A realistic wind turbine simulator is used to validate the achieved performances of the suggested methodology.