the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Data-driven probabilistic surrogate model for floating wind turbine lifetime damage equivalent load prediction
Abstract. Floating offshore wind turbines experience complex hydrodynamic and aerodynamic loading influenced by substructure types and stochastic environmental conditions. Accurately estimating the lifetime fatigue loads requires analyzing thousands of operational scenarios, leading to high computational costs. Moreover, choosing the right input features driving fatigue in floating wind systems and appropriately binning them still remains an open question. We present a fast probabilistic surrogate that maps the site conditions to the loads on the wind turbine. The probabilistic aspect allows the propagation and quantification of statistical uncertainties from the stochastic input quantities on the resulting loads. A fast surrogate eliminates the need to fit a distribution to the site conditions or bin the input data. Rather, all available met-ocean data can be directly used as input, which automatically accounts for the joint distribution in the calculations. The surrogate model in this study uses the mixture density network (MDN) to predict the conditional distribution of the 10-minute damage equivalent loads (DELs) for a 6 MW spar-type floating wind turbine. The MDN achieves high accuracy (R2 > 0.99) in capturing DEL means while efficiently propagating the statistical uncertainties. Furthermore, the surrogate enables quick estimation of 25-year lifetime fatigue damage across a range of potential floating wind farm sites, demonstrating its capability to facilitate rapid decision-making during preliminary site analysis.
- Preprint
(4487 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
CC1: 'Comment on wes-2025-24', Frank Lemmer, 24 Feb 2025
Congratulations to the authors - this is certainly a relevant topic!
I'm wondering if you've seen Kolja Müller's works, which are much related to this work:
Müller, K., & Cheng, P. W. (2018). Application of a Monte Carlo procedure for probabilistic fatigue design of floating offshore wind turbines. Wind Energy Science, 3, 149–162. https://doi.org/10.5194/wes-3-149-2018
Müller, K., & Cheng, P. W. (2018). A surrogate modeling approach for fatigue damage assessment of floating wind turbines. Proceedings of the ASME 37th International Conference on Ocean, Offshore and Arctic Engineering. https://doi.org/10.1115/OMAE2018-78219
Best regardsDisclaimer: this community comment is written by an individual and does not necessarily reflect the opinion of their employer.Citation: https://doi.org/10.5194/wes-2025-24-CC1 - RC1: 'Comment on wes-2025-24', Anonymous Referee #1, 07 Mar 2025
-
RC2: 'Comment on wes-2025-24', Anonymous Referee #2, 27 May 2025
Hi Deepali,
Thank you for this very nice paper :-) The main points to address for the review are:
- Clarify the novelty with respect to your previous paper on Wind Energy Science. It seems that the only change is the database, while the methodology is the same.
- I'm not convinced by the testing set, since it's quite specific instead of random.
- The literature review is well done. Have you found any paper on Bayesian neural networks? Not necessarily from wind. I'm only aware of a report from the HIPERWIND project.
- I had troubles following the hyper-parameters tuning between the main text and the appendix. Some restructuring would be appreciated.
- Which library have you used to train the MDN?
You will find many more comments in the attached pdf. Most of them are just suggestions based on my own experience, while a few are meant to improve the paper. Don't hesitate to get in touch if you need any clarification.
Well done!
Best regards,
Riccardo
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
215 | 58 | 12 | 285 | 18 | 22 |
- HTML: 215
- PDF: 58
- XML: 12
- Total: 285
- BibTeX: 18
- EndNote: 22
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1