Investigation on the use of artificial neural networks to overcome the effects of environmental and operational changes on guided waves monitoring

Intelligent feature extraction and advanced signal processing techniques are necessary for a better interpretation of ultrasonic guided waves signals either in structural health monitoring (SHM) or in nondestructive testing (NDT). Such signals are characterized by at least multi-modal and dispersive components. In addition, in SHM, these signals are closely vulnerable to environmental and operational conditions (EOCs), and can be severely affected. In this paper we investigate the use of Artificial Neural Network (ANN) to overcome these effects and to provide a reliable damage detection method with a minimal of false indications. An experimental case of study (full scale pipe) is presented. Damages sizes have been increased and their shapes modified in different steps. Various parameters such as the number of inputs and the number of hidden neurons were studied to find the optimal configuration of the neural network.

[1]  Hoon Sohn,et al.  Effects of environmental and operational variability on structural health monitoring , 2007, Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[2]  Weaver,et al.  Temperature dependence of diffuse field phase , 1999, Ultrasonics.

[3]  J.E. Michaels,et al.  Detection of structural damage from the local temporal coherence of diffuse ultrasonic signals , 2005, IEEE Transactions on Ultrasonics, Ferroelectrics and Frequency Control.

[4]  Daniel J. Inman,et al.  Impedance-based structural health monitoring of wind turbine blades , 2007, SPIE Smart Structures and Materials + Nondestructive Evaluation and Health Monitoring.

[5]  Charles R. Farrar,et al.  Structural Health Monitoring Using Statistical Pattern Recognition Techniques , 2001 .

[6]  Keith Worden,et al.  An introduction to structural health monitoring , 2007, Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[7]  Piotr Nazarko,et al.  Application of artificial neural networks in the damage identification of structural elements , 2011 .

[8]  Mohammad Bagher Menhaj,et al.  Training feedforward networks with the Marquardt algorithm , 1994, IEEE Trans. Neural Networks.

[9]  Xiao-Hua Yu,et al.  Structural damage detection using artificial neural networks and wavelet transform , 2012, 2012 IEEE International Conference on Computational Intelligence for Measurement Systems and Applications (CIMSA) Proceedings.

[10]  Qian Li,et al.  Optimization of injection molding process parameters using combination of artificial neural network and genetic algorithm method , 2007 .

[11]  Daniel J. Inman,et al.  Structural health monitoring of space rigidizable-inflatable booms , 2006 .

[12]  Babak Moaveni,et al.  Probabilistic Damage Identification of the Dowling Hall Footbridge Using Bayesian FE Model Updating , 2013 .

[13]  Jochen Moll,et al.  Efficient temperature compensation strategies for guided wave structural health monitoring. , 2010, Ultrasonics.

[14]  F. Chang,et al.  Adhesive Layer Effects on PZT-induced Lamb Waves at Elevated Temperatures , 2010 .

[15]  Julian Morris,et al.  Artificial neural networks in process estimation and control , 1992, Autom..