Oil Monitoring and Fault Pre-Warning of Wind Turbine Gearbox Based on Combined Predicting Method

Oil monitoring for wind turbine gearboxes can reflect wear and lubrication conditions, better identify pits on the tooth surface, fatigue wear, other early faults. However, oil with one or several single predicting models brings inaccuracy due to intrinsic merits demerits of models. In this work, fault pre-warning were studied based inspection data three turbines that have been working continuously 3.5 years. The Grey Model (GM) Double Exponential Smoothing (DES) combined by a modified inverse-variance weighting method proposed in which used relative errors calculate weight coefficients, reducing improving accuracy as whole. predicted compared measured verify accuracy. Subsequently, statistical linear regression adopted jointly develop threshold data. Comparing threshold, results showed was warning state. prediction validated an endoscope gearbox, found some parts slightly worn.