|
|
Group of Safety, Reliability and Maintenance > 5.3 System Reliability and Availability > Publications
Ye, HX., Zhu, W.J., Li, H., Ji, W.D., Guedes Soares, C. and Wang, J. (2025), Failure warning for offshore wind turbines based on Autoregressive models, Ocean Engineering, Vol. 332, 121448.
In this paper, a failure warning model for offshore wind turbines is constructed based on the Autoregressive Integrated Moving Average Model (ARIMA), Least Absolute Shrinkage and Selection Operator (LASSO), and multi-condition Exponentially Weighted Moving Average (EWMA) control charts, namely ARIMA-LASSO-EWMA. After variables selection based on correlation analysis, the ARIMA model is developed for each selected variable and the corresponding residuals are computed. The LASSO regression selects features from the residuals obtained by the ARIMA models and extracts the secondary residual as the indicator for the health state recognition. The presence of anomalies in wind turbines is indicated when secondary residuals deviate from the normal range. Consequently, an abnormal detection mechanism based on multi-condition EWMA is established, which sets different warning thresholds for various operating conditions of the wind turbines. Failure warning is issued when the statistic of the secondary residual exceeds the threshold. The results indicate that the proposed failure warning model can provide alarms more than 10 days in advance of actual faults. The proposed model contributes to the condition monitoring, state recognition, and failure warning for offshore wind turbines.
Request a copy of this article
For information about all CENTEC publications you can download: Download the Complete List of CENTEC Publications
|