Determining Remaining Lifetime of Wind Turbine Gearbox Using a Health Status Indicator Signal

Lázaro, R. ; Yildirim Yürüen, N. ; Melero, J.J. (Universidad de Zaragoza)

Resumen: Wind turbine component's failure prognosis allows wind farm owners to apply predictive maintenance techniques to their fleets. This permits optimal scheduling of the maintenance actions considering the best time to stop the turbines and perform those actions. Determining the health status of a turbine's component typically requires verifying a wide number of variables that should be monitored simultaneously. The scope of this study is the investigation and the selection of an effective combination of variables and smoothing and forecasting methodologies for obtaining a wind turbine gearbox health status indicator, in order to interpret clearly the remaining lifetime of the gearbox. The proposed methodology is based on Gaussian Mixture Copula Model (GMCM) models combined with the smoothing treatment and the forecasting model to define the health index of the wind turbine gearbox. Then, the resulting index is tested using various warning and critical thresholds. These thresholds should be chosen adequately in order to indicate appropriate inspection visit and preventive maintenance intervention date. Then, the best combination found, for the studied cases, was 50% and 70% for warning and critical respectively. This combination ensures that the developed procedure is capable of providing long enough time window for maintenance decision making.

Idioma: Inglés

DOI: 10.1088/1742-6596/1618/2/022037

Año: 2020

Publicado en: Journal of Physics: Conference Series 1618, 2 (2020), 022037 [11 pp]
ISSN: 1742-6588

Factor impacto SCIMAGO: 0.21 - Physics and Astronomy (miscellaneous) (Q4)

Tipo y forma: Article (Published version)

Área (Departamento): Área Ingeniería Eléctrica (Dpto. Ingeniería Eléctrica)