Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat
Sign Up
Cited by 60

The DOI matches its Crossref record; 36 of the 36 references it cites were found in a registry or an index.

A Data-Driven Diagnostic Framework for Wind Turbine Structures: A Holistic Approach

Bogoevska, Simona, Spiridonakos, Minas, Chatzi, Eleni, Dumova-Jovanoska, Elena, Höffer, Rudiger·2017·Sensors
DOI 10.3390/s17040720PMC PMC5421680
Open this paper in CoChat — free account

Abstract

The complex dynamics of operational wind turbine (WT) structures challenges the applicability of existing structural health monitoring (SHM) strategies for condition assessment. At the center of Europe’s renewable energy strategic planning, WT systems call for implementation of strategies that may describe the WT behavior in its complete operational spectrum. The framework proposed in this paper relies on the symbiotic treatment of acting environmental/operational variables and the monitored vibration response of the structure. The approach aims at accurate simulation of the temporal variability characterizing the WT dynamics, and subsequently at the tracking of the evolution of this variability in a longer-term horizon. The bi-component analysis tool is applied on long-term data, collected as part of continuous monitoring campaigns on two actual operating WT structures located in different sites in Germany. The obtained data-driven structural models verify the potential of the proposed strategy for development of an automated SHM diagnostic tool.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.3390/s17040720

Cites

36 of 36 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Schematic overview of the proposed bi-component monitoring strategy.
Figure 1 Schematic overview of the proposed bi-component monitoring strategy.CC BY · PubMed Central
Figure 2 Steps of the proposed bi-component monitoring framework.
Figure 2 Steps of the proposed bi-component monitoring framework.CC BY · PubMed Central
Figure 3 Schematic overview of the KF scheme (adapted from [32]).
Figure 3 Schematic overview of the KF scheme (adapted from [32]).CC BY · PubMed Central
Figure 4 Schematic overview of the PCE method: example of two independent input random variables ξ1,ξ2 with uniform pdf and a system output variable Y=ξ1³+cos(ξ2)².
Figure 4 Schematic overview of the PCE method: example of two independent input random variables ξ1,ξ2 with uniform pdf and a system output variable Y=ξ1³+cos(ξ2)².CC BY · PubMed Central
Figure 5 Schematic overview of the experimental setup (case A left and case B right).
Figure 5 Schematic overview of the experimental setup (case A left and case B right).CC BY · PubMed Central
Figure 6 Dynamics of the parked WTs (Case A up and Case B bottom): (a) Stabilization plot for the stationary ARMA and SSI methods (model orders from 2 to 50); (b) Spectrogram and estimated natural frequencies and damping ratios based on an ARMA(32,32) model for Case A and ARMA(18,18) model for Case B (For interpretation of the references to color , the reader is referred to the web version of this article).
Figure 6 Dynamics of the parked WTs (Case A up and Case B bottom): (a) Stabilization plot for the stationary ARMA and SSI methods (model orders from 2 to 50); (b) Spectrogram and estimated natural frequencies and damping ratios based on an ARMA(32,32) model for Case A and ARMA(18,18) model for Case B (For interpretation of the references to color , the reader is referred to the web version of this article).CC BY · PubMed Central
Figure 7 Dynamics of the WTs under normal operation: Stationary ARMA (na,nc) versus nonstationary SP-TARMA (na,nc) model estimates with spectrogram in the background (Short Time Fourier Transform; Hamming data window; NFFT = 512; overlap 98%). (a) Case A: (na,nc) = (32, 32); (b) Case B: (na,nc) = (18, 18).
Figure 7 Dynamics of the WTs under normal operation: Stationary ARMA (na,nc) versus nonstationary SP-TARMA (na,nc) model estimates with spectrogram in the background (Short Time Fourier Transform; Hamming data window; NFFT = 512; overlap 98%). (a) Case A: (na,nc) = (32, 32); (b) Case B: (na,nc) = (18, 18).CC BY · PubMed Central
Figure 8 Statistical criterion for model order selection (Case A top and Case B bottom): (a) Bayesian statistical criterion; (b) Bayesian statistical criterion zoomed-in view; (c) Akaike statistical criterion zoomed-in view (For interpretation of the references to color, the reader is referred to the web version of this article).
Figure 8 Statistical criterion for model order selection (Case A top and Case B bottom): (a) Bayesian statistical criterion; (b) Bayesian statistical criterion zoomed-in view; (c) Akaike statistical criterion zoomed-in view (For interpretation of the references to color, the reader is referred to the web version of this article).CC BY · PubMed Central
Figure 9 SP-TARMA (n = 32, k = 1) model tuning for Case A by varying the value of the ratio of model residuals. First sample of a 10-min acceleration signal fitted with: (a) ν = 0.01; (b) ν = 0.001; (c) ν = 0.0001; Second sample of a 10-min acceleration signal fitted with: (d) ν = 0.01; (e) ν = 0.001; (f) ν = 0.0001.
Figure 9 SP-TARMA (n = 32, k = 1) model tuning for Case A by varying the value of the ratio of model residuals. First sample of a 10-min acceleration signal fitted with: (a) ν = 0.01; (b) ν = 0.001; (c) ν = 0.0001; Second sample of a 10-min acceleration signal fitted with: (d) ν = 0.01; (e) ν = 0.001; (f) ν = 0.0001.CC BY · PubMed Central
Figure 10 SP-TARMA (n = 18, k = 1) model tuning for Case B by varying the value of the ratio of model residuals. First sample of a 10-min acceleration signal fitted with: (a) ν = 0.001; (b) ν = 0.0001; (c) ν = 0.00001; Second sample of a 10-min acceleration signal fitted with: (d) ν = 0.001; (e) ν = 0.0001; (f) ν = 0.00001.
Figure 10 SP-TARMA (n = 18, k = 1) model tuning for Case B by varying the value of the ratio of model residuals. First sample of a 10-min acceleration signal fitted with: (a) ν = 0.001; (b) ν = 0.0001; (c) ν = 0.00001; Second sample of a 10-min acceleration signal fitted with: (d) ν = 0.001; (e) ν = 0.0001; (f) ν = 0.00001.CC BY · PubMed Central
Figure 11 Time history plots of selected PCE input variables for 1388 datasets (Case A).
Figure 11 Time history plots of selected PCE input variables for 1388 datasets (Case A).CC BY · PubMed Central
Figure 12 Scatter plots of selected PCE input variables and estimated correlation values (Case A).
Figure 12 Scatter plots of selected PCE input variables and estimated correlation values (Case A).CC BY · PubMed Central
Figure 13 Input data for Case A: (a) Scatter plots of ICA-based input variables and estimated correlation values, histograms of the latent variables (upper row); (b) Scatter plots of the random vector Ξ of prescribed uniform PDF pΞ(ξ), histograms of the random vector Ξ values (upper row).
Figure 13 Input data for Case A: (a) Scatter plots of ICA-based input variables and estimated correlation values, histograms of the latent variables (upper row); (b) Scatter plots of the random vector Ξ of prescribed uniform PDF pΞ(ξ), histograms of the random vector Ξ values (upper row).CC BY · PubMed Central
Figure 14 (a) Scatter plots of selected PCE input variables and estimated correlation values; (b) Time history plots of the selected PCE input variables for 17 706 datasets (Case B).
Figure 14 (a) Scatter plots of selected PCE input variables and estimated correlation values; (b) Time history plots of the selected PCE input variables for 17 706 datasets (Case B).CC BY · PubMed Central
Figure 15 Input data for Case B: (a) Scatter plots of ICA-based input variables and estimated correlation values, histograms of the latent variables (upper row); (b) Scatter plots of the random vector Ξ of prescribed uniform PDF pΞ(ξ), histograms of the random vector Ξ values (upper row).
Figure 15 Input data for Case B: (a) Scatter plots of ICA-based input variables and estimated correlation values, histograms of the latent variables (upper row); (b) Scatter plots of the random vector Ξ of prescribed uniform PDF pΞ(ξ), histograms of the random vector Ξ values (upper row).CC BY · PubMed Central
Figure 16 PCE estimates for Case A: (a) std of SP-TARMA(32,32,1,0.01) model residual and PCE model expansion values; (b) The PCE errors along with 99.7% confidence intervals (calculated for estimation set).
Figure 16 PCE estimates for Case A: (a) std of SP-TARMA(32,32,1,0.01) model residual and PCE model expansion values; (b) The PCE errors along with 99.7% confidence intervals (calculated for estimation set).CC BY · PubMed Central
Figure 17 PCE estimates for Case B: (a) std of SP-TARMA(18,18,1,0.0001) model residual and PCE model expansion values; (b) The PCE errors along with 99.7% confidence intervals (calculated for estimation set).
Figure 17 PCE estimates for Case B: (a) std of SP-TARMA(18,18,1,0.0001) model residual and PCE model expansion values; (b) The PCE errors along with 99.7% confidence intervals (calculated for estimation set).CC BY · PubMed Central
Figure 18 Time history plots of the selected PCE input variables for 49,956 datasets corresponding to complete year 2012.
Figure 18 Time history plots of the selected PCE input variables for 49,956 datasets corresponding to complete year 2012.CC BY · PubMed Central
Figure 19 Two-month training set: (a) Identified novel data (red points) within time history of 10-min mean values of measured temperature; (b) Identified novel data (red points) within time history of 10-min mean values of measured RPM; (c) Diagnostic Index for the complete data sets of year 2012.
Figure 19 Two-month training set: (a) Identified novel data (red points) within time history of 10-min mean values of measured temperature; (b) Identified novel data (red points) within time history of 10-min mean values of measured RPM; (c) Diagnostic Index for the complete data sets of year 2012.CC BY · PubMed Central
Figure 20 Three-month training set: (a) Identified novel data (red points) within time history of 10-min mean values of measured temperature; (b) Identified novel data (red points) within time history of 10-min mean values of measured RPM; (c) Diagnostic Index for the complete data sets of year 2012.
Figure 20 Three-month training set: (a) Identified novel data (red points) within time history of 10-min mean values of measured temperature; (b) Identified novel data (red points) within time history of 10-min mean values of measured RPM; (c) Diagnostic Index for the complete data sets of year 2012.CC BY · PubMed Central
Figure 21 Four-month training set: (a) Identified novel data (red points) within time history of 10-min mean values of measured temperature; (b) Identified novel data (red points) within time history of 10-min mean values of measured RPM; (c) Diagnostic Index for the complete data sets of year 2012.
Figure 21 Four-month training set: (a) Identified novel data (red points) within time history of 10-min mean values of measured temperature; (b) Identified novel data (red points) within time history of 10-min mean values of measured RPM; (c) Diagnostic Index for the complete data sets of year 2012.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 36 references

Research works printed but not linked: 6

  1. Oliveira, G., Magalhães, F., Cunha, Á., and Caetano, E. (2015, January 19–21). Automated modal tracking and fatigue assessment of a wind turbine based on continuous dynamic monitoring. Proceedings of the 6th International Conference on Experimental Vibration Analysis for Civil Engineering Structures (EVACES 2015), Dubendorf, Switzerland.
  2. Oliveira, G. (2016). Vibration-Based Structural Health Monitoring of Wind Turbines. [Ph.D. Thesis, University of Porto].
  3. Bogoevska, S., Spiridonakos, M., Chatzi, E., Dumova-Jovanoska, E., and Höffer, R. (2016, January 5–8). A novel bi-component structural health monitoring strategy for deriving global models of operational wind turbines. Proceedings of the 8th European Workshop on Structural Health Monitoring (EWSHM 2016), Bilbao, Spain.
  4. Ou, Y., Dertimanis, V., and Chatzi, E. (2016, January 19–21). Experimental Damage Detection of a wind turbine blade under varying operational conditions. Proceedings of the 5th International Conference on Uncertainty in Structural Dynamics (USD 2016), Leuven, Belgium.
  5. Lachmann, S., Baitsch, M., Hartmann, D., and Höffer, R. (2009, January 19–23). Structural lifetime prediction for wind energy converters based on health monitoring and system identification. Proceedings of the 5th European & African Conference on Wind Engineering (EACWE 2009), Florence, Italy.
  6. Oliveira, G., Magalhães, F., Cunha, Á., and Caetano, E. (2016, January 5–8). Optimization of a wind turbine vibration-based SHM system. Proceedings of the 8th European Workshop on Structural Health Monitoring (EWSHM 2016), Bilbao, Spain.

Cited by

See all 60

Metadata

Authors
Bogoevska, S., Spiridonakos, M., Chatzi, E., Dumova-Jovanoska, E., & Höffer, R.
Affiliations
  • Faculty of Civil Engineering, University Ss. Cyril and Methodius, Skopje 1000, Macedonia
  • Department of Civil, Environmental and Geomatic Engineering, ETH, Zürich CH-8093, Switzerland
  • Department of Civil and Environmental Engineering, Ruhr-University Bochum, Bochum 44801, Germany
ORCID
Chatzi, Eleni
Venue
Sensors
Publisher
MDPI AG
Published
2017-03-30
Volume
17
Issue
4
Pages
720
ISSN
1424-8220
DOI
10.3390/s17040720
PMC
PMC5421680
Cited by
60
References
36
Language
en
Licence
creativecommons.org/licenses/by/4.0/ (vor)
Funders
ERC (#67984 WINDMIL)

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: A Data-Driven Diagnostic Framework for Wind Turbine Structures: A Holistic Approach. https://app.cochat.ai/papers/10.3390/s17040720
The paper
Bogoevska, S., Spiridonakos, M., Chatzi, E., Dumova-Jovanoska, E., & Höffer, R. (2017). A Data-Driven Diagnostic Framework for Wind Turbine Structures: A Holistic Approach. Sensors. https://doi.org/10.3390/s17040720

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.