Existing passive acoustics-based techniques for wind turbine blade damage detection lack the robustness and adaptability necessary for an operational implementation due to their physics- and model-based dependency. In contrast, this study develops an entirely unsupervised, data-driven damage detection technique. The novelty of the technique lies in (i) the development and comparison of spectral and cepstral-domain features for the robust characterization of the cavity-internal acoustics, (ii) the use of autoencoder networks to reduce the effects of non-stationary acoustic excitation, and (iii) the exclusion of labeled or damage-case data in the training set. The technique was successfully demonstrated on a wind turbine blade section inflicted with damage of various sizes, types, and locations, and subjected to airflow-induced passive acoustic excitation provided by a wind tunnel. Damage detection accuracy up to 99.82% was achieved for some damage types.
AlíasFSocoróJSevillanoX (2016) A review of physical and perceptual feature extraction techniques for speech, music, and environmental sounds. Applied Sciences6(5): 143.
2.
AnaissiAZandaviSM (2020) Multi-object variational autoencoder: An application for smart infrastructure maintenance. ACM Transactions on Intelligent Systems and Technology9(4): 39.
3.
AntoniadouIDervilisNPapatheouE, et al. (2017) Aspects of structural health and condition monitoring of offshore wind turbines. Philosophical Transactions of The Royal Society A Mathematical Physical and Engineering Sciences373: 20140075.
AtayaSAhmedMMZ (2013) Damages of wind turbine blade trailing edge: Forms, location, and root causes. Engineering Failure Analysis35: 480–488.
6.
AzimiMEslamlouADPekcanG (2020) Data-driven structural health monitoring and damage detection through deep learning: State-of-the-art review. Sensors20: 2778.
7.
BalsamoLBettiR (2015) Data-based structural health monitoring using small training data sets. Structural Control and Health Monitoring22(10): 1240–1264.
8.
BayramBDumanTBInceG (2021) Real time detection of acoustic anomalies in industrial processes using sequential autoencoders. Expert Systems38: e12564.
9.
BealeCInalpolatMNiezreckiC (2020a) Active acoustic damage detection of structural cavities using internal acoustic excitations. Structural Health Monitoring19: 48–65.
10.
BealeCWillisDJNiezreckiC, et al. (2020b) Passive acoustic damage detection of structural cavities using flow-induced acoustic excitations. Structural Health Monitoring19: 751–764.
11.
BergstraJBardenetRBengioY, et al. (2011) Algorithms for Hyper-Parameter Optimization. Advances in Neural Information Processing Systems24.
12.
BerryDSAshwillT (2007) CX-100 manufacturing final project report. Sandia report, SAND2007-6065. Albuquerque, NM: Sandia National Laboratories.
13.
BhargavaVSamalaR (2019) Acoustic emissions from wind turbine blades. Journal of Aerospace Technology and Management11: e4219. DOI: 10.5028/jatm.v11.107.
14.
Bonet-SolàDAlsina-PagèsRM (2021) A comparative survey of feature extraction and machine learning methods in diverse acoustic environments. Sensors21: 1274. DOI: 10.3390/s21041274.
15.
BountourakisVVrysisLKonstantoudakisK, et al. (2019) An enhanced temporal feature integration method for environmental sound recognition. Acoustics1: 410–422.
16.
BouzidOMTianGYCumananK, et al. (2015) Structural health monitoring of wind turbine blades: Acoustic source localization using wireless sensor networks. Journal of Sensors2015: 1–11.
17.
BullLWordenKMansonG, et al. (2018) Active learning for semi-supervised structural health monitoring. Journal of Sound and Vibration437: 373–388.
18.
ChuSNarayananSKuoCCJ (2009) Environmental sound recognition with time–frequency audio features. IEEE Transactions on Audio, Speech, and Language Processing17(6): 1142–1158.
19.
CooperJLeclerqDSteadM (2010) Wind induced aerodynamic noise on microphones from atmospheric measurements. In: Proceedings of the 20th international conference on acoustics, Sydney, Australia, 23–27 August.
20.
DervilisNPapatheouEAntoniadouI, et al. (2016) On the usage of active learning for SHM. In: International conference on noise and vibration engineering, Leuven, Belgium, 19–21 September.
21.
DiezAKhoaNLDMakki AlamdariM, et al. (2016) A clustering approach for structural health monitoring on bridges. Journal of Civil Structural Health Monitoring6(3): 429–445.
22.
DoolanCJMoreauDJBrooksL (2012) Wind turbine noise mechanisms and some concepts for its control. Acoustics Australia40(1): 7–13.
23.
FazendaB (2011) Acoustic based condition monitoring of turbine blades. In: 18th International congress on sound and vibration, Rio de Janeiro, Brazil, 10–14 July.
24.
FigueiredoEParkGFarrarCR, et al. (2011) Machine learning algorithms for damage detection under operational and environmental variability. Structural Health Monitoring10(6): 559–572.
25.
FuruholmHHultbergK (2013) Noise Immission From Wind Farms. Gothenburg, Sweden: Chalmers University of Technology.
26.
GarcíaDTcherniakD (2019) An experimental study on the data-driven structural health monitoring of large wind turbine blades using a single accelerometer and actuator. Mechanical Systems and Signal Processing127: 102–119.
27.
GeorgeJMaryLRiyasKS (2013) Vehicle detection and classification from acoustic signal using ANN and KNN. In: International conference on control, communication, and computing, Thiruvananthapuram, India, 13–15 December.
28.
GoodfelllowIBengioYCourvilleA (2016) Deep Learning. Cambridge, MA: MIT Press.
29.
GriffinD (2019) R&D Pathways for Supersized Wind Turbine Blades. Berkeley, CA: Lawrence Berkeley National Lab.
30.
GulMNecati CatbasF (2009) Statistical pattern recognition for structural health monitoring using time series modeling: Theory and experimental verifications. Mechanical Systems and Signal Processing23: 2192–2204.
31.
HarmanJ (2014) Wind Power in Canada Part 1: Wind Turbine Blades, Gearboxes, and Generators. Toronto, ON: Wind Energy Update.
32.
HubbardHHShephardKP (2009) Wind Turbine Acoustics. Wind Turbine Technology: Fundamental Concepts of Wind Turbine Engineering. Hampton, VA: NASA Langley Research Center, pp.413–465.
33.
InalpolatMNiezreckiC (2020) Acoustic sensing based operational monitoring of wind turbine blades. Journal of Physics Conference Series1452: 012050.
34.
IRENA (2019) Renewable Power Generation Costs in 2018. Abu Dhabi: IRENA.
35.
JiangZHuWDongW, et al. (2018) Structural reliability analysis of wind turbines: A review. Energies10: 2099.
36.
JungYKimYLimH, et al. (2017) Linear scale filterbank for deep neural network based voice activity detection. In: 20th Conference of the oriental chapter of the International coordinating committee on speech databases and speech, Seoul, 1–3 November.
37.
KrauseTOstermannJ (2020) Damage detection for wind turbine rotor blades using airborne sound. Structural Control and Health Monitoring27(5): e2520.
38.
KrauseTPreihsSOstermannJ (2015) Acoustic emission damage detection for wind turbine rotor blades using airborne sound. Structural Health Monitoring. DOI: 10.12783/SHM2015/340.
39.
LeiHLopez-GonzaloE (2009) Mel, linear, and anti-mel frequency cepstral coefficients in broad phonetic regions for telephone speaker recognition. In: 10th International conference of the international speech communication association, Brighton, 6–10 September.
40.
LiDHoSCMSongG, et al. (2015) A review of damage detection methods for wind turbine blades. Smart Materials and Structures24: 033001.
41.
LorenzoEDPetroneGManzatoS, et al. (2016) Damage detection in wind turbine blades by using operational modal analysis. Structural Health Monitoring15(3): 289–301.
42.
MarchiEVesperiniFEybenF, et al. (2015) A novel approach for automatic acoustic novelty detection using a denoising autoencoder with bidirectional LSTM neural networks. In: IEEE international conference on acoustics, speech, and signal processing, South Brisbane, QLD, 19–24 April.
43.
MarquezFPGSanchezPJBRamirezIS (2021) Acoustic inspection system with unmanned aerial vehicles for wind turbines structure health monitoring. Structural Health Monitoring. Epub ahead of print 10April2021. DOI: 10.1177/14759217211004822
44.
Martinez-LuengoMKoliosAWangL (2016) Structural health monitoring of offshore wind turbines: A review through the statistical pattern recognition paradigm. Renewable and Sustainable Energy Reviews64: 91–105.
45.
MATLAB (2021) MATLAB. 2021a Ed. Natick, MA: The Mathworks Inc.
46.
McGuganMLarsenGCSorensenBF, et al. (2008) Fundamentals for Remote Condition Monitoring of Offshore Wind Turbines. Roskilde: Risoe National Laboratory for Sustainable Energy.
MishnaevskyL (2019) Repair of wind turbine blades: Review of methods and related computational mechanics problems. Renewable Energy140: 828–839.
49.
MishnaevskyLBrannerKPetersenHN, et al. (2017) Materials for wind turbine blades: An overview. Materials10(11): 1285.
50.
MisraHIkbalSBourlardH, et al. (2004) Spectral entropy based feature for robust ASR. In: IEEE International conference on acoustics, speech, and signal processing, Montreal, QC, 17–21 May.
51.
MovsessianAGarcía CavaDTcherniakD (2021) An artificial neural network methodology for damage detection: Demonstration on an operating wind turbine blade. Mechanical Systems and Signal Processing159: 107766.
52.
MøllerMF (1993) A scaled conjugate gradient algorithm for fast supervised learning. Neural Networks6(4): 525–533.
53.
NiezreckiCInalpolatM (2015) Structural health monitoring of wind turbine blades using wireless acoustic sensing. University of Massachusetts, Boston, MA, US Patent No. PCT/US2014/062329.
54.
PathirageCSNLiJLiL, et al. (2019) Development and application of a deep learning–based sparse autoencoder framework for structural damage identification. Structural Health Monitoring18: 103–122.
55.
PeetersG (2004) A large set of audio features for sound description (similarity and classification) in the CUIDADO Project. Technical report. IRCAM: Paris, France.
56.
RamachandranRCRamanGDoughertyRP (2014) Wind turbine noise measurement using a compact microphone array with advanced deconvolution algorithms. Journal of Sound and Vibration333(14): 3058–3080.
57.
RandallRB (2017) A history of cepstrum analysis and its application to mechanical problems. Mechanical Systems and Signal Processing97: 3–19.
58.
ReganTBealeCInalpolatM (2017) Wind turbine blade damage detection using supervised machine learning algorithms. Journal of Vibration and Acoustics139: 061010.
59.
RISO/DNV (2002) Guidelines for Design of Wind Turbines, 2nd edn.Roskilde: Wind Energy Department, RISO National Laboratory.
60.
SanchezPJBRamirezISMarquezFPG (2021) Wind Turbines Acoustic Inspections performed with UAV and sound frequency domain analysis. In: 7th International conference on control, instrumentation, and automatation, Tabriz, Iran, 23–24 February.
61.
SerizelRBisotVEssidS, et al. (2017) Acoustic features for environmental sound analysis. In: VirtanenTPlumbleyMDEllisD (eds) Computational Analysis of Sound Scenes and Events, pp.71–101. Springer International Publishing AG.
62.
SolimineJNiezreckiCInalpolatM (2020) An experimental investigation into passive acoustic damage detection for structural health monitoring of wind turbine blades. Structural Health Monitoring19(6): 1711–1725.
63.
SorensenBFJorgensenEDebelCP, et al. (2004) Improved Design of Large Wind Turbine Blade of Fibre Composites Based on Studies of Scale Effects (Phase 1). Roskilde: Risø National Laboratory.
64.
SorensenBFLadingLSendrupP, et al. (2002) Fundamentals for Remote Structural Health Monitoring of Wind Turbine Blades-a Preproject. Roskilde: Risø National Laboratory.
65.
TcherniakDMølgaardLL (2015) Vibration-based SHM system: Application to wind turbine blades. Journal of Physics Conference Series628: 012072.
66.
ThomsenOT (2009) Sandwich materials for wind turbine blades—Present and future. Journal of Sandwich Structures and Materials11: 7–26.
67.
TraylorCDiPaolaMWillisDJ, et al. (2020) A computational investigation of airfoil aeroacoustics for structural health monitoring of wind turbine blades. Wind Energy23(3): 795–809.
68.
TsiapokiSKrauseTHackellM, et al. (2016) Combining a vibration-based SHM scheme and an airborne sound approach for damage detection on wind turbine rotor blades. In: 8th European workshop on structural health monitoring, Bilbao, Spain, 5–8 July.
69.
ValeroXAliasF (2012) Gammatone cepstral coefficients: Biologically inspired features for non-speech audio classification. IEEE Transactions on Multimedia14(6): 1684–1689.
70.
Van Den BergGP (2006) Wind-induced noise in a screened microphone. The Journal of the Acoustical Society of America119(2): 824–833.
71.
YangBSunD (2013) Testing, inspecting, and monitoring technologies for wind turbine blades: A survey. Renewable and Sustainable Energy Reviews22: 515–526.
72.
YingYGarrettJHOppenheimIJ, et al. (2013) Toward data-driven structural health monitoring: Application of machine learning and signal processing to damage detection. Journal of Computing in Civil Engineering27: 667–680.
73.
ZhangGHarichandranRSRamuhalliP (2011) Application of noise cancelling and damage detection algorithms in NDE of concrete bridge decks using impact signals. Journal of Nondestructive Evaluation30: 259–272.
74.
ZhouXGarcia-RomeroDDuraiswamiR, et al. (2011) Linear versus mel frequency cepstral coefficients for speaker recognitio. In: IEEE Workshop on automatic speech recognition and understanding, Waikoloa, HI, 11–15 December.
