Multi-Dimensional Uniform Initialization Gaussian Mixture Model for Spar Crack Quantification under Uncertainty

Guided Wave (GW)-based crack monitoring method as a promising method has been widely studied, as this method is sensitive to small cracks and can cover a wide monitoring range. Online crack quantification is difficult as the initiation and growth of crack are affected by various uncertainties. In addition, crack-sensitive GW features are influenced by time-varying conditions which further increase the difficulty in crack quantification. Considering these uncertainties, the Gaussian mixture model (GMM) is studied to model the probability distribution of GW features. To further improve the accuracy and stability of crack quantification under uncertainties, this paper proposes a multi-dimensional uniform initialization GMM. First, the multi-channel GW features are integrated to increase the accuracy of crack quantification, as GW features from different channels have different sensitivity to cracks. Then, the uniform initialization method is adopted to provide more stable initial parameters in the expectation-maximization algorithm. In addition, the relationship between the probability migration index of GMMs and crack length is calibrated with fatigue tests on prior specimens. Finally, the proposed method is applied for online crack quantification on the notched specimen of an aircraft spar with complex fan-shaped cracks under uncertainty.

[1]  Spilios D. Fassois,et al.  A global statistical model based approach for vibration response-only damage detection under various temperatures: A proof-of-concept study , 2014 .

[2]  Shenfang Yuan,et al.  A Multi-Response-Based Wireless Impact Monitoring Network for Aircraft Composite Structures , 2016, IEEE Transactions on Industrial Electronics.

[3]  Shenfang Yuan,et al.  On-line prognosis of fatigue cracking via a regularized particle filter and guided wave monitoring , 2019, Mechanical Systems and Signal Processing.

[4]  Shenfang Yuan,et al.  Guided Wave-Convolutional Neural Network Based Fatigue Crack Diagnosis of Aircraft Structures , 2019, Sensors.

[5]  Chen Huipeng,et al.  A probabilistic crack size quantification method using in-situ Lamb wave test and Bayesian updating , 2016 .

[6]  M. H. Aliabadi,et al.  Active Health Monitoring of Thick Composite Structures by Embedded and Surface-Mounted Piezo Diagnostic Layer , 2020, Sensors.

[7]  S. Winterstein,et al.  Random Fatigue: From Data to Theory , 1992 .

[8]  Jian Chen,et al.  On-line updating Gaussian process measurement model for crack prognosis using the particle filter , 2020 .

[9]  Anil K. Jain,et al.  Unsupervised Learning of Finite Mixture Models , 2002, IEEE Trans. Pattern Anal. Mach. Intell..

[10]  P. McNicholas.  Mixture Model-Based Classification , 2016 .

[11]  Mira Mitra,et al.  Guided wave based structural health monitoring: A review , 2016 .

[12]  S. Yuan,et al.  A PZT Based On-Line Updated Guided Wave - Gaussian Process Method for Crack Evaluation , 2020, IEEE Sensors Journal.

[13]  Shenfang Yuan,et al.  Damage evaluation by a guided wave-hidden Markov model based method , 2016 .

[14]  C. He,et al.  Obliquely incident EMAT for high-order Lamb wave mode generation based on inclined static magnetic field , 2019, NDT & E International.

[15]  Fu-Kuo Chang,et al.  Encyclopedia of structural health monitoring , 2009 .

[16]  Qiang Wang,et al.  Acousto-ultrasonics-based fatigue damage characterization: Linear versus nonlinear signal features , 2014 .

[17]  J. Celaya,et al.  A multi-feature integration method for fatigue crack detection and crack length estimation in riveted lap joints using Lamb waves , 2013 .

[18]  G. McLachlan,et al.  The EM algorithm and extensions , 1996 .

[19]  Z. Su,et al.  Identification of Damage Using Lamb Waves , 2009 .

[20]  Rui Ding,et al.  A stretchable and large-scale guided wave sensor network for aircraft smart skin of structural health monitoring , 2019, Structural Health Monitoring.

[21]  Antonia Papandreou-Suppappola,et al.  An adaptive learning damage estimation method for structural health monitoring , 2015 .

[22]  Xinlin Qing,et al.  Piezoelectric Transducer-Based Structural Health Monitoring for Aircraft Applications , 2019, Sensors.

[23]  Shiri Gordon,et al.  An efficient image similarity measure based on approximations of KL-divergence between two gaussian mixtures , 2003, Proceedings Ninth IEEE International Conference on Computer Vision.

[24]  Yanfeng Shen,et al.  Nonlinear scattering and mode conversion of Lamb waves at breathing cracks: An efficient numerical approach. , 2019, Ultrasonics.

[25]  Xinlin Qing,et al.  Prediction of Progressive Damage State at the Hot Spots using Statistical Estimation , 2010 .

[26]  Edward Sazonov,et al.  A novel damage index for damage identification using guided waves with application in laminated composites , 2014 .

[27]  Fang Fang,et al.  Improved density peak clustering-based adaptive Gaussian mixture model for damage monitoring in aircraft structures under time-varying conditions , 2019, Mechanical Systems and Signal Processing.