Application of novelty detection methods to health monitoring and typical fault diagnosis of a turbopump

Lei Hu, Niaoqing Hu, Bin Fan, Fengshou Gu

Research output: Contribution to journalConference article

1 Citation (Scopus)

Abstract

Novelty detection is the identification of deviations from a training set. It is suitable for monitoring the health of mechanical systems where it usually is impossible to know every potential fault. In this paper, two novelty detectors are presented. The first detector which integrates One-Class Support Vector Machine (OCSVM) with an incremental clustering algorithm is designed for health monitoring of the turbopump, while the second one which is trained on sensor fault samples is designed to recognize faults from sensors and faults actually from the turbopump. Analysis results showed that these two detectors are both sensitive and efficient for the health monitoring of the turbopump.

Original languageEnglish
Article number012128
Number of pages11
JournalJournal of Physics: Conference Series
Volume364
Issue number1
DOIs
Publication statusPublished - 28 May 2012
Event25th International Congress on Condition Monitoring and Diagnostic Engineering: Sustained Prosperity through Proactive Monitoring, Diagnosis and Management - University of Huddersfield, Huddersfield, United Kingdom
Duration: 18 Jun 201220 Jun 2012
Conference number: 25
http://compeng.hud.ac.uk/comadem2012/ (Link to Conference Website )

Fingerprint

turbine pumps
health
detectors
sensors
education
deviation

Cite this

@article{19cbcc066e9e4ff39615b33382030777,
title = "Application of novelty detection methods to health monitoring and typical fault diagnosis of a turbopump",
abstract = "Novelty detection is the identification of deviations from a training set. It is suitable for monitoring the health of mechanical systems where it usually is impossible to know every potential fault. In this paper, two novelty detectors are presented. The first detector which integrates One-Class Support Vector Machine (OCSVM) with an incremental clustering algorithm is designed for health monitoring of the turbopump, while the second one which is trained on sensor fault samples is designed to recognize faults from sensors and faults actually from the turbopump. Analysis results showed that these two detectors are both sensitive and efficient for the health monitoring of the turbopump.",
author = "Lei Hu and Niaoqing Hu and Bin Fan and Fengshou Gu",
year = "2012",
month = "5",
day = "28",
doi = "10.1088/1742-6596/364/1/012128",
language = "English",
volume = "364",
journal = "Journal of Physics: Conference Series",
issn = "1742-6588",
publisher = "IOP Publishing Ltd.",
number = "1",

}

Application of novelty detection methods to health monitoring and typical fault diagnosis of a turbopump. / Hu, Lei; Hu, Niaoqing; Fan, Bin; Gu, Fengshou.

In: Journal of Physics: Conference Series, Vol. 364, No. 1, 012128, 28.05.2012.

Research output: Contribution to journalConference article

TY - JOUR

T1 - Application of novelty detection methods to health monitoring and typical fault diagnosis of a turbopump

AU - Hu, Lei

AU - Hu, Niaoqing

AU - Fan, Bin

AU - Gu, Fengshou

PY - 2012/5/28

Y1 - 2012/5/28

N2 - Novelty detection is the identification of deviations from a training set. It is suitable for monitoring the health of mechanical systems where it usually is impossible to know every potential fault. In this paper, two novelty detectors are presented. The first detector which integrates One-Class Support Vector Machine (OCSVM) with an incremental clustering algorithm is designed for health monitoring of the turbopump, while the second one which is trained on sensor fault samples is designed to recognize faults from sensors and faults actually from the turbopump. Analysis results showed that these two detectors are both sensitive and efficient for the health monitoring of the turbopump.

AB - Novelty detection is the identification of deviations from a training set. It is suitable for monitoring the health of mechanical systems where it usually is impossible to know every potential fault. In this paper, two novelty detectors are presented. The first detector which integrates One-Class Support Vector Machine (OCSVM) with an incremental clustering algorithm is designed for health monitoring of the turbopump, while the second one which is trained on sensor fault samples is designed to recognize faults from sensors and faults actually from the turbopump. Analysis results showed that these two detectors are both sensitive and efficient for the health monitoring of the turbopump.

UR - http://www.scopus.com/inward/record.url?scp=84862307423&partnerID=8YFLogxK

U2 - 10.1088/1742-6596/364/1/012128

DO - 10.1088/1742-6596/364/1/012128

M3 - Conference article

VL - 364

JO - Journal of Physics: Conference Series

JF - Journal of Physics: Conference Series

SN - 1742-6588

IS - 1

M1 - 012128

ER -