Classification of fatigue damaging segments using artificial neural network

Research output: Contribution to journalArticle

Abstract

This paper focuses on the classification of the fatigue damaging segments datasets associated with the measurement of Variable Amplitude Loadings of strain signals from the coil springs of an automobile during road tests. The wavelet transform was used to extract high damaging segments of the fatigue strain signals. The parameters of the kurtosis, wavelet-based coefficients, and fatigue damage were then calculated for every segment. All the parameters were used as input for the classification analysis using artificial neural networks. Using the back-propagation trained artificial neural network, the corresponding fatigue damages were classified. It was observed that the classification method was able to give 100% accuracy on the classifications based on the damaging segments that were extracted from the training and the validation datasets. From this approach, it classified the level of fatigue damage for coils spring.

Original languageEnglish
Pages (from-to)61-72
Number of pages12
JournalJournal of Mechanical Engineering
Volume5
Issue numberSpecialissue3
Publication statusPublished - 15 Feb 2018

Fingerprint

Fatigue damage
Fatigue of materials
Neural networks
Backpropagation
Wavelet transforms
Automobiles

Keywords

  • Artificial Neural Network
  • Classification
  • Fatigue damaging Segments
  • Variable amplitude loading
  • Wavelet Transform

ASJC Scopus subject areas

  • Mechanical Engineering

Cite this

@article{e239d5cffeeb4dca82bd1e2207868078,
title = "Classification of fatigue damaging segments using artificial neural network",
abstract = "This paper focuses on the classification of the fatigue damaging segments datasets associated with the measurement of Variable Amplitude Loadings of strain signals from the coil springs of an automobile during road tests. The wavelet transform was used to extract high damaging segments of the fatigue strain signals. The parameters of the kurtosis, wavelet-based coefficients, and fatigue damage were then calculated for every segment. All the parameters were used as input for the classification analysis using artificial neural networks. Using the back-propagation trained artificial neural network, the corresponding fatigue damages were classified. It was observed that the classification method was able to give 100{\%} accuracy on the classifications based on the damaging segments that were extracted from the training and the validation datasets. From this approach, it classified the level of fatigue damage for coils spring.",
keywords = "Artificial Neural Network, Classification, Fatigue damaging Segments, Variable amplitude loading, Wavelet Transform",
author = "Yunoh, {M. F.M.} and Shahrir Abdullah and {Md Saad}, {Mohamad Hanif} and {Mohd Nopiah}, Zulkifli and Nuawi, {Mohd. Zaki} and {Mohd Ihsan}, {Ahmad Kamal Ariffin}",
year = "2018",
month = "2",
day = "15",
language = "English",
volume = "5",
pages = "61--72",
journal = "Journal of Mechanical Engineering",
issn = "1823-5514",
publisher = "Institut Pengurusan Penyelidikan (RMI), Universiti Teknologi MARA",
number = "Specialissue3",

}

TY - JOUR

T1 - Classification of fatigue damaging segments using artificial neural network

AU - Yunoh, M. F.M.

AU - Abdullah, Shahrir

AU - Md Saad, Mohamad Hanif

AU - Mohd Nopiah, Zulkifli

AU - Nuawi, Mohd. Zaki

AU - Mohd Ihsan, Ahmad Kamal Ariffin

PY - 2018/2/15

Y1 - 2018/2/15

N2 - This paper focuses on the classification of the fatigue damaging segments datasets associated with the measurement of Variable Amplitude Loadings of strain signals from the coil springs of an automobile during road tests. The wavelet transform was used to extract high damaging segments of the fatigue strain signals. The parameters of the kurtosis, wavelet-based coefficients, and fatigue damage were then calculated for every segment. All the parameters were used as input for the classification analysis using artificial neural networks. Using the back-propagation trained artificial neural network, the corresponding fatigue damages were classified. It was observed that the classification method was able to give 100% accuracy on the classifications based on the damaging segments that were extracted from the training and the validation datasets. From this approach, it classified the level of fatigue damage for coils spring.

AB - This paper focuses on the classification of the fatigue damaging segments datasets associated with the measurement of Variable Amplitude Loadings of strain signals from the coil springs of an automobile during road tests. The wavelet transform was used to extract high damaging segments of the fatigue strain signals. The parameters of the kurtosis, wavelet-based coefficients, and fatigue damage were then calculated for every segment. All the parameters were used as input for the classification analysis using artificial neural networks. Using the back-propagation trained artificial neural network, the corresponding fatigue damages were classified. It was observed that the classification method was able to give 100% accuracy on the classifications based on the damaging segments that were extracted from the training and the validation datasets. From this approach, it classified the level of fatigue damage for coils spring.

KW - Artificial Neural Network

KW - Classification

KW - Fatigue damaging Segments

KW - Variable amplitude loading

KW - Wavelet Transform

UR - http://www.scopus.com/inward/record.url?scp=85052551623&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85052551623&partnerID=8YFLogxK

M3 - Article

VL - 5

SP - 61

EP - 72

JO - Journal of Mechanical Engineering

JF - Journal of Mechanical Engineering

SN - 1823-5514

IS - Specialissue3

ER -