Lamb-wave-based structural health monitoring (SHM) technology for damage location in plate-like structures relies on the postprocessing of captured signals after interacting with damage. Traditional methods typically leverage the time of flight (ToF) of scattered waves from damage. However, these methods are prone to reflected waves from structural boundaries which mix with scattered waves from damage. This is a vital problem faced by most ToF-based detection methods, which seriously narrows the inspection area. To tackle this problem, a machine learning framework, consisting of a multiscale spatiotemporal (MSST) fusion network, is proposed to facilitate the accurate extraction of the ToF of scattered waves through eliminating the influence of boundary reflections. Experiments are conducted with the time-domain Lamb wave signals recorded by a tactically designed piezoelectric sensor array on a 2-mm-thick Al-6061 plate. A pair of circle magnets is attached onto the plate as the wave reflectors. Through step-by-step moving of the magnets in the predefined grids, the corresponding Lamb wave signals are measured to construct a database. An MSST is subsequently designed to minimize the error between estimated and theoretical ToFs, with wavelet coefficients of the signals and transducer position as inputs. The model is trained with the Adam algorithm where 80% of samples in the database are used for training and the rest for evaluation. The final validations are conducted with the scatters off the predefined grids. Results demonstrate that the designed neural network architecture can effectively eliminate boundary reflections and enable precise ToF extraction of the scattered waves from damage. This allows the enlargement of the detection area and presents a promising and useful tool for enhancing the detection performance of existing SHM methods in complex structures.
ChangF-KMarkmillerJFYangJ, et al. Structural health monitoring, system health management. New York, NY: John Wiley and Sons, Inc., 2001.
2.
GiurgiutiuV.Structural health monitoring: with piezoelectric wafer active sensors. Amsterdam, Netherlands: Elsevier, 2007.
3.
FarrarCRWordenK.An introduction to structural health monitoring. Philos Trans A Math Phys Eng Sci2007; 365(1851): 303–315.
4.
AdeagboMOWangS-MNiY-Q.Revamping structural health monitoring of advanced rail transit systems: a paradigmatic shift from digital shadows to digital twins. Adv Eng Inform2024; 61: 102450.
5.
MaleklooAOzerEAlHamaydehM, et al. Machine learning and structural health monitoring overview with emerging technology and high-dimensional data source highlights. Struct Health Monit2022; 21(4): 1906–1955.
6.
SuZYeL.Identification of damage using Lamb waves: from fundamentals to applications. Heidelberg, Germany: Springer Science & Business Media, 2009.
7.
YuLTianZ.Lamb wave structural health monitoring using a hybrid PZT-laser vibrometer approach. Struct Health Monit2013; 12(5–6): 469–483.
8.
KaewniamPCaoMAlkayemNF, et al. Recent advances in damage detection of wind turbine blades: a state-of-the-art review. Renew Sustain Energy Rev2022; 167: 112723.
9.
AnnamdasVGMBhallaSSohCK. Applications of structural health monitoring technology in Asia. Struct Health Monit2017; 16(3): 324–346.
10.
CawleyP.Structural health monitoring: closing the gap between research and industrial deployment. Struct Health Monit2018; 17(5): 1225–1244.
11.
MichaelsJE.Detection, localization and characterization of damage in plates with an in situ array of spatially distributed ultrasonic sensors. Smart Mater Struct2008; 17(3): 035035.
12.
DawsonAJMichaelsJEMichaelsTE.Isolation of ultrasonic scattering by wavefield baseline subtraction. Mech Syst Signal Process2016; 70: 891–903.
13.
HameedMSLiZQiJ, et al. Lamb wave based multistage damage detection method. AIP Conf Proc2019; 2102(1): 050009.
14.
Cantero-ChinchillaSChiachíoJChiachíoM, et al. A robust Bayesian methodology for damage localization in plate-like structures using ultrasonic guided-waves. Mech Syst Signal Process2019; 122: 192–205.
15.
HallJSMcKeonPSatyanarayanL, et al. Minimum variance guided wave imaging in a quasi-isotropic composite plate. Smart Mater Struct2011; 20(2): 025013.
16.
NgC-TVeidtM.A Lamb-wave-based technique for damage detection in composite laminates. Smart Mater Struct2009; 18(7): 074006.
17.
MoriNBiwaSKusakaT.Damage localization method for plates based on the time reversal of the mode-converted Lamb waves. Ultrasonics2019; 91: 19–29.
18.
QiuLYuanS.A phase synthesis time reversal impact imaging method for on-line composite structure monitoring. Smart Struct Syst2011; 8(3): 303–320.
19.
PurekarAPinesD.Damage detection in thin composite laminates using piezoelectric phased sensor arrays and guided Lamb wave interrogation. J Intell Mater Syst Struct2010; 21(10): 995–1010.
20.
YuWShenfangYLeiQ.Improved wavelet-based spatial filter of damage imaging method on composite structures. Chinese J Aeronaut2011; 24(5): 665–672.
21.
BinLLeiQShenfangY.Damage imaging and localization method based on multi-dimension arrays and spatial filter without wave velocity. Acta Materize Compositae Sinica2014; 31(3): 835–844.
22.
QiuLLiuBYuanS, et al. Impact imaging of aircraft composite structure based on a model-independent spatial-wavenumber filter. Ultrasonics2016; 64: 10–24.
23.
RenYQiuLYuanS, et al. A diagnostic imaging approach for online characterization of multi-impact in aircraft composite structures based on a scanning spatial-wavenumber filter of guided wave. Mech Syst Signal Process2017; 90: 44–63.
24.
EngholmMStepinskiT.Direction of arrival estimation of Lamb waves using circular arrays. Struct Health Monit2011; 10(5): 467–480.
25.
ZuoHYangZXuC, et al. Damage identification for plate-like structures using ultrasonic guided wave based on improved MUSIC method. Compos Struct2018; 203: 164–171.
26.
BaoQYuanSWangY, et al. Anisotropy compensated MUSIC algorithm based composite structure damage imaging method. Compos Struct2019; 214: 293–303.
27.
SalmanpourMSharif KhodaeiZAliabadiM.Transducer placement optimisation scheme for a delay and sum damage detection algorithm. Struct Control Health Monit2017; 24(4): e1898.
BaddooPJHerrmannBMcKeonBJ, et al. Physics-informed dynamic mode decomposition. Proc R Soc A2023; 479(2271): 20220576.
31.
WuYSicardBGadsdenSAJESwA. Physics-informed machine learning: a comprehensive review on applications in anomaly detection and condition monitoring. Expert Syst Appl2024; 255: 124678.
32.
ShuklaKDi LeoniPCBlackshireJ, et al. Physics-informed neural network for ultrasound nondestructive quantification of surface breaking cracks. J Nondestruct Eval2020; 39: 1–20.
33.
FarrarCRWordenK.Structural health monitoring: a machine learning perspective. Hoboken, NJ: John Wiley & Sons, 2012.
34.
BaoYLiH.Machine learning paradigm for structural health monitoring. Struct Health Monitor2021; 20(4): 1353–1372.
35.
ZhangSLiCMYeW.Damage localization in plate-like structures using time-varying feature and one-dimensional convolutional neural network. Mech Syst Signal Process2021; 147: 107107.
36.
ZhangSLiCMYangJ, et al. Effective combination of modeling and experimental data with deep metric learning for guided wave-based damage localization in plates. Mech Syst Signal Process2022; 172: 108979.
37.
WangXLinMLiJ, et al. Ultrasonic guided wave imaging with deep learning: applications in corrosion mapping. Mech Syst Signal Process2022; 169: 108761.
38.
RautelaMGopalakrishnanA. Deep learning frameworks for wave propagation-based damage detection in 1d-waveguides. In: Proceedings of 11th international symposium on NDT in aerospace, Toulouse, France, 23–26 September 2019, vol. 2, pp. 1–11.
39.
SbarufattiCMansonGWordenK.A numerically-enhanced machine learning approach to damage diagnosis using a Lamb wave sensing network. J Sound Vib2014; 333(19): 4499–4525.
40.
LinMGuoSHeS, et al. Structure health monitoring of a composite wing based on flight load and strain data using deep learning method. Compos Struct2022; 286: 115305.
41.
HeeschMDziendzikowskiMMendrokK, et al. Diagnostic-quality guided wave signals synthesized using generative adversarial neural networks. Sensors2022; 22(10): 3848.
42.
LiuCChenYXuXJ.Structural digital twin for damage detection of CFRP composites using meta transfer learning-based approach. Expert Syst Appl2024; 261: 125458.
43.
XuXLiuCJ.Physics-guided deep learning for damage detection in CFRP composite structures. Compos Struct2024; 331: 117889.
44.
LiuXJiangZYanZ.Improvement of accuracy in damage localization using frequency slice wavelet transform. Shock Vib2012; 19(4): 585–596.
45.
AmariS-i.Backpropagation and stochastic gradient descent method. Neurocomputing1993; 5(4–5): 185–196.
46.
ZhangZ. Improved adam optimizer for deep neural networks. In: 2018 IEEE/ACM 26th international symposium on quality of service (IWQoS), Banff, AB, Canada, 04–06 June 2018. Piscataway, NJ: IEEE, 2018, pp. 1–2.
47.
MaQChenELinZ, et al. Convolutional multitimescale echo state network. IEEE Trans Cybern2019; 51(3): 1613–1625.
48.
NairVHintonGE.Rectified linear units improve restricted boltzmann machines. In: Proceedings of the 27th international conference on machine learning (ICML-10), Haifa, Israel, 21–24 June2010, pp. 807–814. Madison, WI: Omnipress.
49.
IoffeSSzegedyC.Batch normalization: accelerating deep network training by reducing internal covariate shift. In: International conference on machine learning. Lille, France, 6–11 July 2015, Cambridge, MA: PMLR, 2015, pp. 448–456.
50.
GholamalinezhadHKhosraviH.Pooling methods in deep neural networks, a review. arXiv preprint arXiv:2009.07485, 2020.
51.
KruseRMostaghimSBorgeltC, et al. Multi-layer perceptrons. In: Computational intelligence: a methodological introduction. Cham, Switzerland: Springer, 2022, pp. 53–124.
52.
MontgomeryDCPeckEAViningGG.Introduction to linear regression analysis. Hoboken, NJ: John Wiley & Sons, 2021.
SongY-YYingLJ.Decision tree methods: applications for classification and prediction. Shanghai Arch Psychiatry2015; 27(2): 130.
56.
MaćkiewiczARatajczakWJC. Principal components analysis (PCA). Comput Geosci1993; 19(3): 303–342.
57.
AwadMKhannaRAwadM, et al. Efficient learning machines: theories, concepts, and applications for engineers and system designers. New York, NY: Apress, 2015, pp. 67–80.
58.
LeeDSeungHS.Algorithms for non-negative matrix factorization. Adv Neural Inf Process Syst2000; 13: 556–562.
59.
SherstinskyAJ.Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network. Phys D Nonlinear Phenom2020; 404: 132306.
ChungJGulcehreCChoK, et al. Empirical evaluation of gated recurrent neural networks on sequence modeling. In: Proceedings of the NIPS workshop on deep learning, Montreal, Canada, 8–13 December 2014. Cambridge, MA: NeurIPS Foundation.
VaswaniA.Attention is all you need. Adv Neural Inform Process Syst2017.
64.
HuangGLiuZVan Der MaatenL, et al. Densely connected convolutional networks. In: Proceedings of the IEEE conference on computer vision and pattern recognition, Honolulu, Hawaii, 21–26 July, 2017, pp. 4700–4708. Piscataway, NJ: IEEE publisher.
65.
HeKZhangXRenS, et al. Deep residual learning for image recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, Las Vegas, Nevada, 27–30 June 2016, pp. 770–778. Piscataway, NJ: IEEE publisher.
66.
KrizhevskyASutskeverIHintonGE.ImageNet classification with deep convolutional neural networks. Adv Neural Inform Process Syst2012; 25: 1097–1105.
67.
BelkinaACCiccolellaCOAnnoR, et al. Automated optimized parameters for T-distributed stochastic neighbor embedding improve visualization and analysis of large datasets. Nat Commun2019; 10(1): 5415.
68.
ChenYWangYChenY, et al. Deep autoencoder for interpretable tissue-adaptive deconvolution and cell-type-specific gene analysis. Nat Commun2022; 13(1): 6735.
69.
MaYXuZ-QJZhangJ.Frequency principle in deep learning beyond gradient-descent-based training. arXiv preprint arXiv:2101.00747, 2021.
70.
ZhengRWangGHoK.Accurate semidefinite relaxation method for elliptic localization with unknown transmitter position. IEEE Trans Wirel Commun2020; 20(4): 2746–2760.
