Comparison of machine learning methods for crack localization

In this paper, the Haar wavelet discrete transform, the artificial neural networks (ANNs), and the random forests (RFs) are applied to predict the location and severity of a crack in an Euler–Bernoulli cantilever subjected to the transverse free vibration. An extensive investigation into two data collection sets and machine learning methods showed that the depth of a crack is more difficult to predict than its location. The data set of eight natural frequency parameters produces more accurate predictions on the crack depth; meanwhile, the data set of eight Haar wavelet coefficients produces more precise predictions on the crack location. Furthermore, the analysis of the results showed that the ensemble of 50 ANN trained by Bayesian regularization and Levenberg–Marquardt algorithms slightly outperforms RF.

[1]  S. Caddemi,et al.  Exact solution of the multi-cracked Euler–Bernoulli column , 2008 .

[2]  M. Attar.  A transfer matrix method for free vibration analysis and crack identification of stepped beams with multiple edge cracks and different boundary conditions , 2012 .

[3]  Marta B. Rosales,et al.  Crack detection in beam-like structures , 2009 .

[4]  Ekrem Tufekci,et al.  Exact solution based finite element formulation of cracked beams for crack detection , 2016 .

[5]  T. Chondros,et al.  Analytical Methods in Rotor Dynamics , 1983 .

[6]  Andrew D. Dimarogonas,et al.  Vibration of cracked beams , 1989 .

[7]  Leontios J. Hadjileontiadis,et al.  Fractal dimension analysis for crack identification in beam structures , 2005 .

[8]  Nikos A. Aspragathos,et al.  Identification of crack location and magnitude in a cantilever beam from the vibration modes , 1990 .

[9]  Danilo Iglesias Brandão,et al.  Electronic Instrumentation for Distributed Generation and Power Processes , 2017 .

[10]  Alice E. Smith,et al.  Performance measures, consistency, and power for artificial neural network models , 1995 .

[11]  Krzysztof Wilde,et al.  Application of continuous wavelet transform in vibration based damage detection method for beams and plates , 2006 .

[12]  J. Kurnitski,et al.  New higher order Haar wavelet method: Application to FGM structures , 2018, Composite Structures.

[13]  Marsha Mitchell-Blanks,et al.  Performance Measures , 2009, Encyclopedia of Database Systems.

[14]  P. Konar,et al.  Bearing fault detection of induction motor using wavelet and Support Vector Machines (SVMs) , 2011, Appl. Soft Comput..

[15]  Mohamed A. Mahmoud,et al.  Neural network solution of the inverse vibration problem , 1999 .

[16]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[17]  F. Chu,et al.  Using genetic algorithms and finite element methods to detect shaft crack for rotor-bearing system , 2001 .

[18]  M. Rucka,et al.  Crack identification using wavelets on experimental static deflection profiles , 2006 .

[19]  H. Abdul Razak,et al.  Fault diagnosis on beam-like structures from modal parameters using artificial neural networks , 2015 .

[20]  B. P. Nandwana,et al.  Modelling of vibration of beam in presence of inclined edge or internal crack for its possible detection based on frequency measurements , 1997 .

[21]  James Hensman,et al.  Natural computing for mechanical systems research: A tutorial overview , 2011 .

[22]  S. Caddemi,et al.  Multi-cracked Euler-Bernoulli beams: Mathematical modeling and exact solutions , 2013 .

[23]  Y. Yana,et al.  Development in vibration-based structural damage detection technique , 2007 .

[24]  Mihir Kumar Sutar,et al.  Neural Based Controller for Smart Detection of Crack in Cracked Cantilever Beam , 2015 .