Preprints
https://doi.org/10.5194/wes-2024-160
https://doi.org/10.5194/wes-2024-160
27 Nov 2024
 | 27 Nov 2024
Status: this preprint is currently under review for the journal WES.

Gaussian Mixture autoencoder for uncertainty-aware damage identification in a Floating Offshore Wind Turbine

Ana Fernandez-Navamuel, Nicolas Gorostidi, David Pardo, Vincenzo Nava, and Eleni Chatzi

Abstract. This work proposes an uncertainty-aware approach to the inverse problem of damage identification in a Floating Offshore Wind Turbine (FOWT). We design an autoencoder architecture, where the latent space represents the features of the target damaged condition. The inverse operator (encoder) is a Deep Neural Network that maps the measurable response to the parameters (means, variances, and weights) of a multivariate Gaussian Mixture model. The Gaussian Mixture model provides a convenient distributional description that is flexible enough to accommodate complex solution spaces. The decoder receives samples from the Gaussian Mixture and maps the damaged condition (states) to the system’s measurable response. In such a problem, and depending on the quantities being observed (sensor positioning), it is possible that multiple damaged states may correspond to similar measurement records. In this context, the main contribution of this work lies in the development of a method to quantify the uncertainty within the context of a possibly ill-posed damage identification problem. We employ the Gaussian Mixture to express the multimodal solution space and explain the uncertainty in the damaged condition estimates. We design and validate the methodology using synthetic data from a FOWT in the commonly adopted OpenFAST software, and consider two damage types frequently occurring in mooring lines: biofouling and anchor displacement. The method allows for estimating the damaged state while capturing the uncertainty in the estimations and the multimodality of the solution under the availability of a limited number of response measurements.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.
Ana Fernandez-Navamuel, Nicolas Gorostidi, David Pardo, Vincenzo Nava, and Eleni Chatzi

Status: open (until 28 Dec 2024)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
Ana Fernandez-Navamuel, Nicolas Gorostidi, David Pardo, Vincenzo Nava, and Eleni Chatzi
Ana Fernandez-Navamuel, Nicolas Gorostidi, David Pardo, Vincenzo Nava, and Eleni Chatzi

Viewed

Total article views: 106 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
73 29 4 106 2 3
  • HTML: 73
  • PDF: 29
  • XML: 4
  • Total: 106
  • BibTeX: 2
  • EndNote: 3
Views and downloads (calculated since 27 Nov 2024)
Cumulative views and downloads (calculated since 27 Nov 2024)

Viewed (geographical distribution)

Total article views: 106 (including HTML, PDF, and XML) Thereof 106 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 08 Dec 2024
Download
Short summary
This work employs Deep Neural Networks to identify damage in the mooring system of a Floating Offshore Wind Turbine (FOWT) using measurements from the platform response. We account for the effect of uncertainty in the assessment estimates caused by the existence of multiple solutions (different damage scenarios can cause the observed data). We describe the damage condition features using a distributional model based on a Gaussian Mixture, which captures the uncertainty in the predictions.
Altmetrics