Signal model-based fault detection and diagnosis for induction motors using features of vibration signal in two-dimension domain

Van Tuan Do, Ui-Pil Chong

Research output: Contribution to journalArticleScientificpeer-review

48 Citations (Scopus)

Abstract

In this paper, we propose an approach for vibration signal-based fault detection and diagnosis system applying for induction motors. The approach consists of two consecutive processes: fault detection process and fault diagnosis process. In the fault detection process, significant features from vibration signals are extracted through the scale invariant feature transform (SIFT) algorithm to generate the faulty symptoms. Consequently, the pattern classification technique using the faulty symptoms is applied to the fault diagnosis process. Hence, instead of analyzing the vibration signal to determine the induction motor faults, the vibration signal can be classified to the corresponding faulty category, which presents the induction motor fault. We also provide a framework for the pattern classification technique that is applicable to SIFT patterns. Moreover, a comparison with two other approaches in our previous work is also carried out. The testing results show that our proposed approach provides significantly high fault classification accuracy and a better performance than previous approaches.
Original languageEnglish
Pages (from-to)655-666
JournalStrojniški vestnik
Volume57
Issue number9
DOIs
Publication statusPublished - 2011
MoE publication typeA1 Journal article-refereed

Fingerprint

Fault detection
Induction motors
Failure analysis
Pattern recognition
Testing

Keywords

  • Classification accuracy
  • fault detection and diagnosis
  • feature vector
  • SIFT
  • texton dictionary
  • two-dimension domain

Cite this

@article{6edbd3f90c7d48e58cdb9ca43eccbbc2,
title = "Signal model-based fault detection and diagnosis for induction motors using features of vibration signal in two-dimension domain",
abstract = "In this paper, we propose an approach for vibration signal-based fault detection and diagnosis system applying for induction motors. The approach consists of two consecutive processes: fault detection process and fault diagnosis process. In the fault detection process, significant features from vibration signals are extracted through the scale invariant feature transform (SIFT) algorithm to generate the faulty symptoms. Consequently, the pattern classification technique using the faulty symptoms is applied to the fault diagnosis process. Hence, instead of analyzing the vibration signal to determine the induction motor faults, the vibration signal can be classified to the corresponding faulty category, which presents the induction motor fault. We also provide a framework for the pattern classification technique that is applicable to SIFT patterns. Moreover, a comparison with two other approaches in our previous work is also carried out. The testing results show that our proposed approach provides significantly high fault classification accuracy and a better performance than previous approaches.",
keywords = "Classification accuracy, fault detection and diagnosis, feature vector, SIFT, texton dictionary, two-dimension domain",
author = "Do, {Van Tuan} and Ui-Pil Chong",
year = "2011",
doi = "10.5545/sv-jme.2010.162",
language = "English",
volume = "57",
pages = "655--666",
journal = "Strojniški vestnik",
issn = "0039-2480",
publisher = "Assoc. of Mechanical Eng. and Technicians of Slovenia",
number = "9",

}

Signal model-based fault detection and diagnosis for induction motors using features of vibration signal in two-dimension domain. / Do, Van Tuan; Chong, Ui-Pil.

In: Strojniški vestnik, Vol. 57, No. 9, 2011, p. 655-666.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Signal model-based fault detection and diagnosis for induction motors using features of vibration signal in two-dimension domain

AU - Do, Van Tuan

AU - Chong, Ui-Pil

PY - 2011

Y1 - 2011

N2 - In this paper, we propose an approach for vibration signal-based fault detection and diagnosis system applying for induction motors. The approach consists of two consecutive processes: fault detection process and fault diagnosis process. In the fault detection process, significant features from vibration signals are extracted through the scale invariant feature transform (SIFT) algorithm to generate the faulty symptoms. Consequently, the pattern classification technique using the faulty symptoms is applied to the fault diagnosis process. Hence, instead of analyzing the vibration signal to determine the induction motor faults, the vibration signal can be classified to the corresponding faulty category, which presents the induction motor fault. We also provide a framework for the pattern classification technique that is applicable to SIFT patterns. Moreover, a comparison with two other approaches in our previous work is also carried out. The testing results show that our proposed approach provides significantly high fault classification accuracy and a better performance than previous approaches.

AB - In this paper, we propose an approach for vibration signal-based fault detection and diagnosis system applying for induction motors. The approach consists of two consecutive processes: fault detection process and fault diagnosis process. In the fault detection process, significant features from vibration signals are extracted through the scale invariant feature transform (SIFT) algorithm to generate the faulty symptoms. Consequently, the pattern classification technique using the faulty symptoms is applied to the fault diagnosis process. Hence, instead of analyzing the vibration signal to determine the induction motor faults, the vibration signal can be classified to the corresponding faulty category, which presents the induction motor fault. We also provide a framework for the pattern classification technique that is applicable to SIFT patterns. Moreover, a comparison with two other approaches in our previous work is also carried out. The testing results show that our proposed approach provides significantly high fault classification accuracy and a better performance than previous approaches.

KW - Classification accuracy

KW - fault detection and diagnosis

KW - feature vector

KW - SIFT

KW - texton dictionary

KW - two-dimension domain

U2 - 10.5545/sv-jme.2010.162

DO - 10.5545/sv-jme.2010.162

M3 - Article

VL - 57

SP - 655

EP - 666

JO - Strojniški vestnik

JF - Strojniški vestnik

SN - 0039-2480

IS - 9

ER -