Neural networks are commonly recognized tools for the classification of multidimensional data obtained in structural health monitoring (SHM) systems. Their configuration for a given scenario is, however, a challenging task, which limits the possibilities of their practical applications. In this article the authors propose using the neural network ensemble approach for the classification of SHM data generated by guided wave sensor networks. The overproduce and choose strategy is used for designing ensembles containing different types and sizes of neural networks. The proposed method allows for a significant increase of the state assessment reliability, which is illustrated by the results obtained from the practical industrial case of a full-scale aircraft test. The method is verified in the process of detecting fatigue cracks propagating in the aircraft load-carrying structure. The long-term experiments are performed in variable environmental conditions with a net of structure-embedded piezoelectric sensors.
BaraldiPCompareMSaucoS. (2013) Ensemble neural network-based particle filtering for prognostics. Mechanical Systems and Signal Processing41(1–2): 288–300.
2.
BishopCM (1995) Neural Networks for Pattern Recognition. Oxford: Clarendon Press.
3.
DaoPBStaszewskiWJ (2013) Data normalisation for Lamb wave-based damage detection using cointegration: A case study with single-and multiple-temperature trends. Journal of Intelligent Material Systems and Structures25(7): 1045389X13512186.
4.
DraganKŚwiderskiW (2010) Studying efficiency of NDE techniques applied to composite materials in aerospace applications. Acta Physica Polonica Series A117(5): 878.
5.
DworakowskiZAmbrozinskiLDraganK. (2014) Voting Neural network classifier for detection of fatigue damage in aircrafts. In: 7th European workshop on structural health monitoring, La Cite, Nantes, France, 8–11 July 2014, pp.1894–1901. France: Inria.
6.
DworakowskiZAmbrozinskiLPackoP. (2015) Application of artificial neural networks for compounding multiple damage indices in Lamb-wave-based damage detection. Structural Control and Health Monitoring22: 50–61.
GiacintoGRoliF (2001) An approach to the automatic design of multiple classifier systems. Pattern Recognition Letters22: 25–33.
9.
HansenLSalamonP (1990) Neural network ensembles. IEEE Transactions on Pattern Analysis and Machine Intelligence12(10): 993–1001.
10.
HaykinS (2001) Neural Networks A Comprehensive Foundation. Pearson Prentice Hall.
11.
KijankaPManoharALanza di ScaleaF. (2014) Damage location by ultrasonic Lamb waves and piezoelectric rosettes. Journal of Intelligent Material Systems and Structures26(12): 1477–1490.
KolakowskiP (2006) Two approaches to structural damage identification: model updating versus soft computing. Journal of Intelligent Material Systems and Structures17(1): 63–79.
LiZYanXTianZ. (2013) Blind vibration component separation and nonlinear feature extraction applied to the nonstationary vibration signals for the gearbox multi-fault diagnosis. Measurement: Journal of the International Measurement Confederation46: 259–271.
16.
LiewCKVeidtM (2009) Pattern recognition of guided waves for damage evaluation in bars. Pattern Recognition Letters30(3): 321–330.
17.
LuYYeLSuZ. (2009) Artificial neural network (ANN)-based crack identification in aluminum plates with Lamb wave signals. Journal of Intelligent Material Systems and Structures20(1): 39–49.
18.
MendrokK (2014) Multiple damage localization using local modal filters. Diagnostyka15(3): 15–21.
19.
NgCT (2014) On the selection of advanced signal processing techniques for guided wave damage identification using a statistical approach. Engineering Structures67: 50–60.
20.
PerroneMCooperL (1993) When networks disagree: Ensemble methods for hybrid neural networks. In: MammoneRJ (ed.) Neural Networks for Speech and Image Processing. Boca Raton, FL: Chapman and Hall, pp.126–142.
21.
PieczonkaL (2014) The influence of excitation frequency on the effectiveness of vibrothermographic testing. Diagnostyka15(4): 47–51.
22.
PuthillathPRoseJL (2010) Ultrasonic guided wave inspection of a titanium repair patch bonded to an aluminum aircraft skin. International Journal of Adhesion and Adhesives30(7): 566–573.
23.
QiuLLiuBYuanS (2015) Impact imaging of aircraft composite structure based on a model-independent spatial-wavenumber filter. Ultrasonics64: 10–24.
24.
QiuLYuanSChangFK. (2014) On-line updating Gaussian mixture model for aircraft wing spar damage evaluation under time-varying boundary condition. Smart Materials and Structures23(12): 125001.
25.
SalamoneSBartoliILanza Di ScaleaF. (2009) Guided-wave Health monitoring of aircraft composite panels under changing temperature. Journal of Intelligent Material Systems and Structures20: 1079–1090.
26.
SbarufattiCMansonGWordenK (2014) A numerically-enhanced machine learning approach to damage diagnosis using a Lamb wave sensing network. Journal of Sound and Vibration333(19): 1–27.
27.
SesmeroMPAlonso-weberJMGutierrezG. (2015) An ensemble approach of dual base learners for multi-class classification problems. Information Fusion24: 122–136.
28.
SharkeyAJCSharkeyNEGereckeU. (2000) The “test and select” approach to ensemble combination. In: KittlerJRoliF (eds) Multiple Classifier Systems. Berlin, Heidelberg: Springer-Verlag, pp.30–44.
29.
SuZYeL (2009) Identification of Damage Using Lamb Waves, From Fundamentals to Applications. London, Germany: Springer-Verlag.
30.
UhlTKohutPHolakK. (2011) Vision based condition assessment of structures. Journal of Physics: Conference Series305: 012043.
31.
WordenKStaszewskiWJHensmanJJ (2011) Natural computing for mechanical systems research: A tutorial overview. Mechanical Systems and Signal Processing25(1): 4–111.
32.
YuanSWangLPengG (2005) Neural network method based on a new damage signature for structural health monitoring. Thin-Walled Structures43: 553–563.
