An evaluation of the potential offered by a relevance vector classifier in machinery fault diagnosis

Zhang Kui, Li Yuhua, Fan Yibo, Gu Fengshou, Andrew Ball

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

The increasing complexity of modern machinery systems demands an effective fault diagnosis strategy with low cost, high efficiency and reliability. This paper reports work which attempts to explore the potential offered by a Relevance Vector Machine (RVM) in machinery fault diagnosis. This work starts with a full investigation into the demands of modern fault diagnosis and the characteristics of the RVM method, and then provides an insight into the model of a relevance vector machine for classification. Finally, a case study of the multi-class classification of bearing faults further demonstrates the application potential of the method. Besides, it is proved that the proposed method is most suitable for real-time applications due to its high computational speed, low memory requirement and high accuracy.

Original languageEnglish
Pages (from-to)35-40
Number of pages6
JournalInternational Journal of COMADEM
Volume9
Issue number4
Publication statusPublished - 1 Oct 2006
Externally publishedYes

Fingerprint

Failure analysis
Machinery
Classifiers
Bearings (structural)
Data storage equipment
Classifier
Evaluation
Fault diagnosis
Costs
Demand systems
Fault

Cite this

@article{f930d4851b8f4b54a0d5859890e033db,
title = "An evaluation of the potential offered by a relevance vector classifier in machinery fault diagnosis",
abstract = "The increasing complexity of modern machinery systems demands an effective fault diagnosis strategy with low cost, high efficiency and reliability. This paper reports work which attempts to explore the potential offered by a Relevance Vector Machine (RVM) in machinery fault diagnosis. This work starts with a full investigation into the demands of modern fault diagnosis and the characteristics of the RVM method, and then provides an insight into the model of a relevance vector machine for classification. Finally, a case study of the multi-class classification of bearing faults further demonstrates the application potential of the method. Besides, it is proved that the proposed method is most suitable for real-time applications due to its high computational speed, low memory requirement and high accuracy.",
keywords = "Fault diagnosis, Pattern classification, Relevance vector machine, Support vector machine",
author = "Zhang Kui and Li Yuhua and Fan Yibo and Gu Fengshou and Andrew Ball",
year = "2006",
month = "10",
day = "1",
language = "English",
volume = "9",
pages = "35--40",
journal = "International Journal of COMADEM",
issn = "1363-7681",
publisher = "COMADEM International",
number = "4",

}

An evaluation of the potential offered by a relevance vector classifier in machinery fault diagnosis. / Kui, Zhang; Yuhua, Li; Yibo, Fan; Fengshou, Gu; Ball, Andrew.

In: International Journal of COMADEM, Vol. 9, No. 4, 01.10.2006, p. 35-40.

Research output: Contribution to journalArticle

TY - JOUR

T1 - An evaluation of the potential offered by a relevance vector classifier in machinery fault diagnosis

AU - Kui, Zhang

AU - Yuhua, Li

AU - Yibo, Fan

AU - Fengshou, Gu

AU - Ball, Andrew

PY - 2006/10/1

Y1 - 2006/10/1

N2 - The increasing complexity of modern machinery systems demands an effective fault diagnosis strategy with low cost, high efficiency and reliability. This paper reports work which attempts to explore the potential offered by a Relevance Vector Machine (RVM) in machinery fault diagnosis. This work starts with a full investigation into the demands of modern fault diagnosis and the characteristics of the RVM method, and then provides an insight into the model of a relevance vector machine for classification. Finally, a case study of the multi-class classification of bearing faults further demonstrates the application potential of the method. Besides, it is proved that the proposed method is most suitable for real-time applications due to its high computational speed, low memory requirement and high accuracy.

AB - The increasing complexity of modern machinery systems demands an effective fault diagnosis strategy with low cost, high efficiency and reliability. This paper reports work which attempts to explore the potential offered by a Relevance Vector Machine (RVM) in machinery fault diagnosis. This work starts with a full investigation into the demands of modern fault diagnosis and the characteristics of the RVM method, and then provides an insight into the model of a relevance vector machine for classification. Finally, a case study of the multi-class classification of bearing faults further demonstrates the application potential of the method. Besides, it is proved that the proposed method is most suitable for real-time applications due to its high computational speed, low memory requirement and high accuracy.

KW - Fault diagnosis

KW - Pattern classification

KW - Relevance vector machine

KW - Support vector machine

UR - http://www.scopus.com/inward/record.url?scp=33947367403&partnerID=8YFLogxK

UR - http://usir.salford.ac.uk/33131/

UR - http://www.comadem.com/journal/

M3 - Article

VL - 9

SP - 35

EP - 40

JO - International Journal of COMADEM

JF - International Journal of COMADEM

SN - 1363-7681

IS - 4

ER -