Research Papers

Wind Turbine Blade Damage Detection Using Supervised Machine Learning Algorithms

[+] Author and Article Information
Taylor Regan, Christopher Beale

Department of Mechanical Engineering,
University of Massachusetts Lowell,
Lowell, MA 01854

Murat Inalpolat

Department of Mechanical Engineering,
University of Massachusetts Lowell,
1 University Avenue,
Lowell, MA 01854
e-mail: Murat_Inalpolat@uml.edu

1Corresponding author.

Contributed by the Technical Committee on Vibration and Sound of ASME for publication in the JOURNAL OF VIBRATION AND ACOUSTICS. Manuscript received June 10, 2016; final manuscript received May 26, 2017; published online August 2, 2017. Assoc. Editor: Patrick S. Keogh.

J. Vib. Acoust 139(6), 061010 (Aug 02, 2017) (14 pages) Paper No: VIB-16-1291; doi: 10.1115/1.4036951 History: Received June 10, 2016; Revised May 26, 2017

Wind turbine blades undergo high operational loads, experience variable environmental conditions, and are susceptible to failure due to defects, fatigue, and weather-induced damage. These large-scale composite structures are fundamentally enclosed acoustic cavities and currently have limited, if any, structural health monitoring (SHM) in place. A novel acoustics-based structural sensing and health monitoring technique is developed, requiring efficient algorithms for operational damage detection of cavity structures. This paper describes the selection of a set of statistical features for acoustics-based damage detection of enclosed cavities, such as wind turbine blades, as well as a systematic approach used in the identification of competent machine learning algorithms. Logistic regression (LR) and support vector machine (SVM) methods are identified and used with optimal feature selection for decision-making via binary classification algorithms. A laboratory-scale wind turbine with hollow composite blades was built for damage detection studies. This test rig allows for testing of stationary or rotating blades, of which time and frequency domain information can be collected to establish baseline characteristics. The test rig can then be used to observe any deviations from the baseline characteristics. An external microphone attached to the tower will be utilized to monitor blade health while blades are internally ensonified by wireless speakers. An initial test campaign with healthy and damaged blade specimens is carried out to arrive at several conclusions on the detectability and feature extraction capabilities required for damage detection.

Copyright © 2017 by ASME
Your Session has timed out. Please sign back in to continue.


AWEA, 2015, “ U.S. Wind Industry Market Reports,” American Wind Energy Association, Washington, DC, accessed June 9, 2017, http://www.awea.org/2015-market-reports
DOE, 2015, “ Enabling Wind Power Nationwide,” U.S. Department of Energy, Washington, DC, Technical Report No. DOE/EE-12187068. https://www.osti.gov/scitech/biblio/1220457-enabling-wind-power-nationwide
Lantz, E. , Wiser, R. , and Hand, M. , 2012, “ The Past and the Future Cost of Wind Energy,” National Renewable Energy Laboratory, Golden, CO, Report No. NREL/CP-6A20-54526. http://www.nrel.gov/docs/fy12osti/54526.pdf
Regan, T. , Canturk, R. , Slavkovsky, E. , Niezrecki, C. , and Inalpolat, M. , “ Wind Turbine Blade Damage Detection Using Various Machine Learning Algorithms,” ASME Paper No. DETC2016-59686.
Arora, V. , Wijnant, Y. H. , and de Boer, A. , 2014, “ Acoustic-Based Damage Detection Method,” Appl. Acoust., 80, pp. 23–27. [CrossRef]
Yang, W. , 2013, “ Testing and Condition Monitoring of Composite Wind Turbine Blades,” Recent Advances in Composite Materials for Wind Turbine Blade, World Academic Publishing, Hong Kong, pp. 147–169.
Hyers, R. W. , McGowan, J. G. , Sullivan, K. L. , Manwell, J. F. , and Syrett, B. C. , 2006, “ Condition Monitoring and Prognosis of Utility Scale Wind Turbines,” Energy Mater., 1(3), pp. 187–203. [CrossRef]
Kahn-Jetter, Z. L. , and Chu, T. C. , 1990, “ Three-Dimensional Displacement Measurements Using Digital Image Correlation and Photogrammic Analysis,” Exp. Mech., 30(1), pp. 10–16. [CrossRef]
Ozbek, M. , Rixen, D. J. , Erne, O. , and Sanow, G. , 2010, “ Feasibility of Monitoring Large Wind Turbines Using Photogrammetry,” Energy, 35(12), pp. 4802–4811. [CrossRef]
Niezrecki, C. , Avitable, P. , Chen, J. , Sherwood, J. , Lundstrom, T. , LeBlanc, B. , Hughes, S. , Desmond, M. , Beattie, A. , Rumsey, M. , Klute, S. M. , Pedrazzani, R. , Werlink, R. , and Newman, J. , 2014, “ Inspection and Monitoring of Wind Turbine Blade-Embedded Wave Defects During Fatigue Testing,” Struct. Health Monit., 13(6), pp. 1–15. [CrossRef]
Tipperman, J. , and Lanza di Scalea, F. , 2014, “ Experiments on a Wind Turbine Blade Testing an Indication for Damage Using the Causal and Anti-Causal Green's Function Reconstructed From a Diffuse Field,” Proc. SPIE, 9064, p. 90641-1.
Boukabache, H. , Escriba, C. , Zedek, S. , and Fourniols, J. , 2013, “ Wavelet Decomposition Based Diagnostic for Structural Health Monitoring on Metallic Aircrafts: Case of Crack Triangulation and Corrosion Detection,” Int. J. Prognostics Health Manage., 4(3), pp. 1–9. http://www.phmsociety.org/sites/phmsociety.org/files/phm_submission/2012/ijphm_13_003.pdf
Aizawa, K. , and Niezrecki, C. , 2014, “ Wind Turbine Blade Health Monitoring Using Acoustic Beamforming Techniques,” J. Acoust. Soc. Am., 135(4), pp. 2392–2393.
Aizawa, K. , Poozesh, P. , Niezrecki, C. , Baqersad, J. , Inalpolat, M. , and Heilmann, G. , 2015, “ An Acoustic-Array Based Structural Health Monitoring Technique for Wind Turbine Blades,” Proc. SPIE, 9437, p. 94371P.
Fazenda, B. M. , 2011, “ Acoustic Based Condition Monitoring of Turbine Blades,” 18th International Congress on Sound and Vibration (ICSV), Rio de Janeiro, Brazil, July 10–14, pp. 1–8. http://usir.salford.ac.uk/15862/
Fazenda, B. M. , and Comboni, D. , 2012, “ Acoustic Condition Monitoring of Wind Turbines: Tip Faults,” The Ninth International Conference on Condition Monitoring and Machinery Failure Prevention Technologies (CM and MFPT), London, June 12–14, pp. 1–15. http://usir.salford.ac.uk/27344/1/Fazenda_Comboni_BINDT2012_Final_Paper.pdf
Stearman, R. O. , Schulz, G. H. , and Rohre, S. M. , 1997, “ Aircraft Damage Detection From Acoustic and Noise Impressed Signals Found by a Cockpit Voice Recorder,” J. Acoust. Soc. Am., 101(5), p. 3085. [CrossRef]
Lam, H. F. , Ng, C. T. , Lee, Y. Y. , and Sun, H. Y. , 2009, “ System Identification of an Enclosure With Leakages Using a Probabilistic Approach,” J. Sound Vib., 322(4–5), pp. 756–771. [CrossRef]
Nair, K. K. , and Kiremidjian, A. S. , 2006, “ Time Series Based Structural Damage Detection Algorithm Using Gaussian Mixtures Modeling,” ASME J. Dyn. Syst. Meas. Control, 129(3), pp. 285–293. [CrossRef]
Sohn, H. , Farrar, C. R. , Hemez, F. M. , Shunk, D. D. , Stinemates, D. W. , and Nadler, B. R. , 2003, “ A Review of Structural Health Monitoring Literature From 1996–2001,” Los Alamos National Laboratory, Los Alamos, NM, Report No. LA-13976-MS. http://citeseerx.ist.psu.edu/viewdoc/download;jsessionid=43453048A5725D068F53552AE67579DB?doi=
Dervilis, N. , Shi, H. , Worden, K. , and Cross, E. J. , 2016, “ Exploring Environmental and Operational Variations in SHM Data Using Heteroscedastic Gaussian Processes,” Dyn. Civ. Struct., 2, pp. 145–153.
Krause, T. , Preihs, S. , and Ostermann, J. , 2014, “ Detection of Impulse-Like Airborne Sound for Damage Identification in Rotor Blades of Wind Turbines,” Seventh European Workshop on Structural Health Monitoring (EWSHM), Nantes, France, July 8–11, pp. 488–495. https://tel.archives-ouvertes.fr/EWSHM-TUBT6/hal-01020385v1
Edwards, H. , Neal, K. , Reilly, J. , and Van Buren, K. , 2016, “ Making Structural Condition Diagnostics Robust to Environmental Variability,” Dynamics of Civil Structures (Vol. 2, Conference Proceedings of the Society for Experimental Mechanics Series), Springer, Cham, Switzerland, pp. 117–130. [CrossRef]
Loh, C. , and Chan, C. , 2015, “ Damage Assessment of Steel Structures Using Multi-Autoregressive Model,” Dyn. Civ. Struct., 2, pp. 1–8.
Figueiredo, E. , Park, G. , Farrar, C. R. , Worden, K. , and Figueiras, J. , 2010, “ Machine Learning Algorithms for Damage Detection Under Operational and Environmental Variability,” Struct. Health Monit., 10(6), pp. 559–572. [CrossRef]
Nick, W. , Shelton, J. , Asamene, K. , and Esterline, A. , 2015, “ A Study of Supervised Machine Learning Techniques for Structural Health Monitoring,” 26th Modern AI and Cognitive Science Conference (MAICS), Greensboro, NC, Apr. 25–26, Vol. 1353, Paper No. 36. http://ceur-ws.org/Vol-1353/paper_36.pdf
Worden, K. , and Manson, G. , 2007, “ The Application of Machine Learning to Structural Health Monitoring,” Philos. Trans. R. Soc., A, 365(1851), pp. 515–537. [CrossRef]
Niezrecki, C. , and Inalpolat, M. , 2015, “ Structural Health Monitoring of Wind Turbine Blades Using Wireless Acoustic Sensing,” University of Massachusetts, Boston, MA, U.S. Patent No. PCT/US2014/062329. https://patentscope.wipo.int/search/en/detail.jsf?docId=WO2015065873
Canturk, R. , and Inalpolat, M. , 2016, “Development of an Acoustic Sensing Based SHM Technique for Wind Turbine Blades,” International Modal Analysis Conference (IMAC0XXXIV), Orlando, FL, Jan. 25–28, pp. 95–104.
Canturk, R. , and Inalpolat, M. , 2015, “ A Computational Acoustic Interrogation of Wind Turbine Blades With Damage,” Comsol Conference, Boston, MA, Oct. 7–9, pp. 1–6. https://www.comsol.com/paper/a-computational-acoustic-interrogation-of-damage-to-wind-turbine-blades-26192
Vachtsevanos, G. , Lewis, F. , Roemer, M. , Hess, A. , and Wu, B. , 2006, Intelligent Fault Diagnosis and Prognosis for Engineering Systems, Wiley, Hoboken, NJ. [CrossRef]
Farrar, C. R. , and Worden, K. , 2007, “ An Introduction to Structural Health Monitoring,” Philos. Trans. R. Soc., A, 365(1851), pp. 303–315.
Cheng, S. , Azarian, M. H. , and Pecht, M. G. , 2010, “ Sensor Systems for Prognostics and Health Management,” Sensors, 10(6), pp. 5774–5797. [CrossRef] [PubMed]
Farrar, C. R. , and Worden, K. , 2013, Structural Health Monitoring: A Machine Learning Perspective, Wiley, Hoboken, NJ.
Le, Q. V. , Ngiam, J. , Coates, A. , Lahiri, A. , Prochnow, B. , and Ng, A. , 2011, “ On Optimization Methods for Deep Learning,” 28th International Conference on Machine Learning (ICML), Bellevue, WA, June 28–July 2, pp. 1–8. http://ai.stanford.edu/~quocle/LeNgiCoaLahProNg11.pdf
MathWorks, 2014, “ MATLAB R2014,” MathWorks Inc., Natick, MA.
Hastie, T. , Tibshirani, R. , and Friedman, J. , 2009, The Elements of Statistical Learning, Springer-Verlag, New York. [CrossRef]


Grahic Jump Location
Fig. 3

General process overview for the supervised learning algorithms

Grahic Jump Location
Fig. 2

Illustration of the feature distinguishability metric

Grahic Jump Location
Fig. 1

Schematic of the active damage detection

Grahic Jump Location
Fig. 4

The sigmoid hypothesis function, hθ(x)=g(z)

Grahic Jump Location
Fig. 11

Schematic of blade 1 damage locations

Grahic Jump Location
Fig. 5

Logistic regression cost curves for y = 1 and y = 0

Grahic Jump Location
Fig. 12

Instances of accuracies greater than 98%

Grahic Jump Location
Fig. 6

Multiple hyperplanes fitting sample dataset

Grahic Jump Location
Fig. 7

Illustration of the steps in hyperplane formulation

Grahic Jump Location
Fig. 8

Ill-fitting and optimal hyperplane comparison

Grahic Jump Location
Fig. 10

(a) Solid model of the subscale turbine and (b) completed subscale turbine prototype

Grahic Jump Location
Fig. 9

Cost functions for (a) y = 1 and (b) y = 0

Grahic Jump Location
Fig. 15

Comparison of the peak amplitude FFT and mean frequency features for a stationary multi-mid excitation test

Grahic Jump Location
Fig. 16

Comparison of the RMS and kurtosis features for a stationary multi-high excitation test

Grahic Jump Location
Fig. 13

Combined damage case testing accuracies

Grahic Jump Location
Fig. 14

Machine learning accuracies for rotating tests

Grahic Jump Location
Fig. 17

The two-feature plot for all test cases from a stationary multi-mid excitation test

Grahic Jump Location
Fig. 21

Multi-high tip hole distinguishability results

Grahic Jump Location
Fig. 20

Distinguishability distribution for (a) case 1, (b) case 2, and (c) case 3

Grahic Jump Location
Fig. 18

A side-by-side comparison of a feature pair with and without the first level of damage included

Grahic Jump Location
Fig. 19

A side-by-side two-feature plot comparing the variability of healthy data clusters across test cases

Grahic Jump Location
Fig. 22

Fisher's ratio for all tests and excitation types



Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In