DAMAGE ASSESSMENT OF BEAMS USING AN ARTIFICIAL NEURAL NETWORK AND NATURAL FREQUENCIES.

Saved in:

Bibliographic Details
Title:	DAMAGE ASSESSMENT OF BEAMS USING AN ARTIFICIAL NEURAL NETWORK AND NATURAL FREQUENCIES.
Authors:	Tufisi, Cristian¹, Gillich, Gilbert-Rainer¹, Popescu, Cristinel², Ardeljan, Mario¹
Source:	Annals of 'Constantin Brancusi' University of Targu-Jiu. Engineering Series / Analele Universităţii Constantin Brâncuşi din Târgu-Jiu. Seria Inginerie. 2021, Issue 2, p44-51. 8p.
Subjects:	Artificial neural networks, Feedforward neural networks, Bayesian analysis
Abstract:	The current paper presents a modal-based damage identification method that uses a one-layer neural network based on Bayesian regularization to estimate the location and severity of transverse cracks present in cantilever beams. A feedforward neural network is created, and it is trained by employing the relative frequency shift curves (RFS) for known damage positions and depths. The RFS values are plotted by using the squared modal curvature for the first eight weak-axis vibration modes and an enhanced method for assessing the severity for cracks of different depths. Training data were obtained for specific damage positions, by removing the transverse crack with a step of 10 mm along the cantilever beam. The ANN system is trained both to depict the location and severity of transverse cracks present in cantilever beams, and the results obtained are evaluated using data generated from FEM analysis. [ABSTRACT FROM AUTHOR]
	Copyright of Annals of 'Constantin Brancusi' University of Targu-Jiu. Engineering Series / Analele Universităţii Constantin Brâncuşi din Târgu-Jiu. Seria Inginerie is the property of Universitatea Constantin Brancusi din Targu-Jiu and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	The current paper presents a modal-based damage identification method that uses a one-layer neural network based on Bayesian regularization to estimate the location and severity of transverse cracks present in cantilever beams. A feedforward neural network is created, and it is trained by employing the relative frequency shift curves (RFS) for known damage positions and depths. The RFS values are plotted by using the squared modal curvature for the first eight weak-axis vibration modes and an enhanced method for assessing the severity for cracks of different depths. Training data were obtained for specific damage positions, by removing the transverse crack with a step of 10 mm along the cantilever beam. The ANN system is trained both to depict the location and severity of transverse cracks present in cantilever beams, and the results obtained are evaluated using data generated from FEM analysis. [ABSTRACT FROM AUTHOR]
ISSN:	18424856