Articles | Volume 10, issue 9
https://doi.org/10.5194/wes-10-2099-2025
https://doi.org/10.5194/wes-10-2099-2025
Research article
 | 
29 Sep 2025
Research article |  | 29 Sep 2025

In situ condition monitoring of floating offshorewind turbines using kurtosis anddeep-learning-based approaches

Adrien Hirvoas, Cesar Aguilera, Matthieu Perrault, Damien Desbordes, and Romain Ribault

Related authors

Hydro-elastic coupling effect on the dynamic global response of a spar-type floating offshore wind turbine
Cesar Aguilera, Romain Ribault, Jerome De-Lauzon, and Adrien Hirvoas
Wind Energ. Sci. Discuss., https://doi.org/10.5194/wes-2025-110,https://doi.org/10.5194/wes-2025-110, 2025
Preprint under review for WES
Short summary

Cited articles

Bangalore, P. and Tjernberg, L. B.: An artificial neural network approach for early fault detection of gearbox bearings, IEEE T. Smart Grid, 6, 980–987, 2015. a
Barahona, B., Hoelzl, C., and Chatzi, E.: Applying design knowledge and machine learning to scada data for classification of wind turbine operating regimes, in: 2017 IEEE Symposium Series on Computational Intelligence (SSCI), Honolulu, HI, USA, 2017, 1–8, https://doi.org/10.1109/SSCI.2017.8280947, 2017. a
Berahmand, K., Daneshfar, F., Salehi, E. S., Li, Y., and Xu, Y.: Autoencoders and their applications in machine learning: a survey, Artif. Intell. Rev., 57, 28, https://doi.org/10.1007/s10462-023-10662-6, 2024. a, b, c
Brincker, R. and Ventura, C.: Introduction to operational modal analysis, John Wiley & Sons, ISBN 978-1-119-96315-8, 2015. a
Chen, J., Li, J., Chen, W., Wang, Y., and Jiang, T.: Anomaly detection for wind turbines based on the reconstruction of condition parameters using stacked denoising autoencoders, Renew. Energ., 147, 1469–1480, 2020. a
Download
Short summary
This study compares two methods for detecting downtime in the Zefyros floating offshore wind turbine off Norway's coast. The first method uses kurtosis to find signal anomalies, while the second employs a deep-learning autoencoder to reduce and reconstruct data, identifying irregularities. Using 1 month of sensor data, the study finds the deep-learning method to be more accurate than kurtosis. This research advances the use of deep learning for effective wind turbine monitoring.
Share
Altmetrics
Final-revised paper
Preprint