17 Citations (Scopus)

Abstract

This paper presents the use of a kernel-based machine learning technique, popular in the field of pattern recognition, to detect and classify various forms of damage states in both isotropic and anisotropic structures. A classification algorithm based on one-class Support Vector Machines (SVMs) is used for damage detection. The SVMs use a Gaussian kernel to map the input attributes to the high dimensional feature space and the transformed features are linearly separated by a decision plane. A procedure for obtaining the optimal value of the hyperparameter that controls the smoothness of the kernel is described. The type of damage addressed in this paper includes a combination of loose bolt and fatigue crack damage in a single lap, Al 6061-T651, bolted joints. Graphite/epoxy composite laminates with different types of damage are also studied, taking into account uncertainties in the measurement and material properties. The results show that the algorithm is able to accurately distinguish between different torque states and changes in crack length in the bolted joint sample. In anisotropic media, the algorithm was able to detect and classify various damage signatures with significant accuracy, using mutual information of two sensors. The algorithm was able to produce similar levels of accuracy when variability due to material properties was introduced.

Original languageEnglish (US)
Pages (from-to)451-458
Number of pages8
JournalInsight: Non-Destructive Testing and Condition Monitoring
Volume49
Issue number8
DOIs
StatePublished - Aug 2007

Fingerprint

damage
Bolted joints
bolted joints
Support vector machines
Materials properties
Graphite epoxy composites
Anisotropic media
Damage detection
Bolts
cracks
graphite-epoxy composites
Pattern recognition
Laminates
Learning systems
bolts
machine learning
Torque
anisotropic media
Cracks
pattern recognition

Keywords

  • Anomaly detection
  • Bolted joint
  • Classification
  • Fatigue crack
  • Kernel
  • Structural health monitoring
  • Support Vector Machines
  • Time embedding technique

ASJC Scopus subject areas

  • Instrumentation
  • Materials Science (miscellaneous)

Cite this

Damage diagnosis using a kernel-based method. / Chattopadhyay, Aditi; Das, Santanu; Coelho, Clyde K.

In: Insight: Non-Destructive Testing and Condition Monitoring, Vol. 49, No. 8, 08.2007, p. 451-458.

Research output: Contribution to journalArticle

@article{6ef5ebedb19e41c983b363bf8543eab1,
title = "Damage diagnosis using a kernel-based method",
abstract = "This paper presents the use of a kernel-based machine learning technique, popular in the field of pattern recognition, to detect and classify various forms of damage states in both isotropic and anisotropic structures. A classification algorithm based on one-class Support Vector Machines (SVMs) is used for damage detection. The SVMs use a Gaussian kernel to map the input attributes to the high dimensional feature space and the transformed features are linearly separated by a decision plane. A procedure for obtaining the optimal value of the hyperparameter that controls the smoothness of the kernel is described. The type of damage addressed in this paper includes a combination of loose bolt and fatigue crack damage in a single lap, Al 6061-T651, bolted joints. Graphite/epoxy composite laminates with different types of damage are also studied, taking into account uncertainties in the measurement and material properties. The results show that the algorithm is able to accurately distinguish between different torque states and changes in crack length in the bolted joint sample. In anisotropic media, the algorithm was able to detect and classify various damage signatures with significant accuracy, using mutual information of two sensors. The algorithm was able to produce similar levels of accuracy when variability due to material properties was introduced.",
keywords = "Anomaly detection, Bolted joint, Classification, Fatigue crack, Kernel, Structural health monitoring, Support Vector Machines, Time embedding technique",
author = "Aditi Chattopadhyay and Santanu Das and Coelho, {Clyde K.}",
year = "2007",
month = "8",
doi = "10.1784/insi.2007.49.8.451",
language = "English (US)",
volume = "49",
pages = "451--458",
journal = "Insight: Non-Destructive Testing and Condition Monitoring",
issn = "1354-2575",
publisher = "British Institute of Non-Destructive Testing",
number = "8",

}

TY - JOUR

T1 - Damage diagnosis using a kernel-based method

AU - Chattopadhyay, Aditi

AU - Das, Santanu

AU - Coelho, Clyde K.

PY - 2007/8

Y1 - 2007/8

N2 - This paper presents the use of a kernel-based machine learning technique, popular in the field of pattern recognition, to detect and classify various forms of damage states in both isotropic and anisotropic structures. A classification algorithm based on one-class Support Vector Machines (SVMs) is used for damage detection. The SVMs use a Gaussian kernel to map the input attributes to the high dimensional feature space and the transformed features are linearly separated by a decision plane. A procedure for obtaining the optimal value of the hyperparameter that controls the smoothness of the kernel is described. The type of damage addressed in this paper includes a combination of loose bolt and fatigue crack damage in a single lap, Al 6061-T651, bolted joints. Graphite/epoxy composite laminates with different types of damage are also studied, taking into account uncertainties in the measurement and material properties. The results show that the algorithm is able to accurately distinguish between different torque states and changes in crack length in the bolted joint sample. In anisotropic media, the algorithm was able to detect and classify various damage signatures with significant accuracy, using mutual information of two sensors. The algorithm was able to produce similar levels of accuracy when variability due to material properties was introduced.

AB - This paper presents the use of a kernel-based machine learning technique, popular in the field of pattern recognition, to detect and classify various forms of damage states in both isotropic and anisotropic structures. A classification algorithm based on one-class Support Vector Machines (SVMs) is used for damage detection. The SVMs use a Gaussian kernel to map the input attributes to the high dimensional feature space and the transformed features are linearly separated by a decision plane. A procedure for obtaining the optimal value of the hyperparameter that controls the smoothness of the kernel is described. The type of damage addressed in this paper includes a combination of loose bolt and fatigue crack damage in a single lap, Al 6061-T651, bolted joints. Graphite/epoxy composite laminates with different types of damage are also studied, taking into account uncertainties in the measurement and material properties. The results show that the algorithm is able to accurately distinguish between different torque states and changes in crack length in the bolted joint sample. In anisotropic media, the algorithm was able to detect and classify various damage signatures with significant accuracy, using mutual information of two sensors. The algorithm was able to produce similar levels of accuracy when variability due to material properties was introduced.

KW - Anomaly detection

KW - Bolted joint

KW - Classification

KW - Fatigue crack

KW - Kernel

KW - Structural health monitoring

KW - Support Vector Machines

KW - Time embedding technique

UR - http://www.scopus.com/inward/record.url?scp=34548224574&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34548224574&partnerID=8YFLogxK

U2 - 10.1784/insi.2007.49.8.451

DO - 10.1784/insi.2007.49.8.451

M3 - Article

AN - SCOPUS:34548224574

VL - 49

SP - 451

EP - 458

JO - Insight: Non-Destructive Testing and Condition Monitoring

JF - Insight: Non-Destructive Testing and Condition Monitoring

SN - 1354-2575

IS - 8

ER -