Selection of Input Parameters for Multivariate Classifiers in Proactive Machine Health Monitoring by Clustering Envelope Spectrum Harmonics

Research output: Contribution to journalArticle

Abstract

In condition monitoring (CM) signal analysis the inherent problem of key characteristics being masked by noise can be addressed by analysis of the signal envelope. Envelope analysis of vibration signals is effective in extracting useful information for diagnosing different faults. However, the number of envelope features is generally too large to be effectively incorporated in system models. In this paper a novel method of extracting the pertinent information from such signals based on multivariate statistical techniques is developed which substantialy reduces the number of input parameters required for data classification models. This was achieved by clustering possible model variables into a number of homogeneous groups to assertain levels of interdependency. Representatives from each of the groups were selected for their power to discriminate between the categorical classes. The techniques established were applied to a reciprocating compressor rig wherein the target was identifying machine states with respect to operational health through comparison of signal outputs for healthy and faulty systems. The technique allowed near perfect fault classification. In addition methods for identifying seperable classes are investigated through profiling techniques, illustrated using Andrew’s Fourier curves.
LanguageEnglish
Pages308-313
Number of pages6
JournalApplied Mechanics and Materials
Volume798
DOIs
Publication statusPublished - Oct 2015

Fingerprint

Classifiers
Health
Monitoring
Reciprocating compressors
Signal analysis
Condition monitoring

Cite this

@article{093b30e3c4dd48a4b595237cb31b5904,
title = "Selection of Input Parameters for Multivariate Classifiers in Proactive Machine Health Monitoring by Clustering Envelope Spectrum Harmonics",
abstract = "In condition monitoring (CM) signal analysis the inherent problem of key characteristics being masked by noise can be addressed by analysis of the signal envelope. Envelope analysis of vibration signals is effective in extracting useful information for diagnosing different faults. However, the number of envelope features is generally too large to be effectively incorporated in system models. In this paper a novel method of extracting the pertinent information from such signals based on multivariate statistical techniques is developed which substantialy reduces the number of input parameters required for data classification models. This was achieved by clustering possible model variables into a number of homogeneous groups to assertain levels of interdependency. Representatives from each of the groups were selected for their power to discriminate between the categorical classes. The techniques established were applied to a reciprocating compressor rig wherein the target was identifying machine states with respect to operational health through comparison of signal outputs for healthy and faulty systems. The technique allowed near perfect fault classification. In addition methods for identifying seperable classes are investigated through profiling techniques, illustrated using Andrew’s Fourier curves.",
keywords = "classifiers, Cluster analysis, Genetic algorithms, reciprocating compressor, Relevance Vector Machine",
author = "Ann Smith and Andrew Ball and Fengshou Gu",
year = "2015",
month = "10",
doi = "10.4028/www.scientific.net/AMM.798",
language = "English",
volume = "798",
pages = "308--313",
journal = "Applied Mechanics and Materials",
issn = "1662-7482",
publisher = "Trans Tech Publications",

}

TY - JOUR

T1 - Selection of Input Parameters for Multivariate Classifiers in Proactive Machine Health Monitoring by Clustering Envelope Spectrum Harmonics

AU - Smith, Ann

AU - Ball, Andrew

AU - Gu, Fengshou

PY - 2015/10

Y1 - 2015/10

N2 - In condition monitoring (CM) signal analysis the inherent problem of key characteristics being masked by noise can be addressed by analysis of the signal envelope. Envelope analysis of vibration signals is effective in extracting useful information for diagnosing different faults. However, the number of envelope features is generally too large to be effectively incorporated in system models. In this paper a novel method of extracting the pertinent information from such signals based on multivariate statistical techniques is developed which substantialy reduces the number of input parameters required for data classification models. This was achieved by clustering possible model variables into a number of homogeneous groups to assertain levels of interdependency. Representatives from each of the groups were selected for their power to discriminate between the categorical classes. The techniques established were applied to a reciprocating compressor rig wherein the target was identifying machine states with respect to operational health through comparison of signal outputs for healthy and faulty systems. The technique allowed near perfect fault classification. In addition methods for identifying seperable classes are investigated through profiling techniques, illustrated using Andrew’s Fourier curves.

AB - In condition monitoring (CM) signal analysis the inherent problem of key characteristics being masked by noise can be addressed by analysis of the signal envelope. Envelope analysis of vibration signals is effective in extracting useful information for diagnosing different faults. However, the number of envelope features is generally too large to be effectively incorporated in system models. In this paper a novel method of extracting the pertinent information from such signals based on multivariate statistical techniques is developed which substantialy reduces the number of input parameters required for data classification models. This was achieved by clustering possible model variables into a number of homogeneous groups to assertain levels of interdependency. Representatives from each of the groups were selected for their power to discriminate between the categorical classes. The techniques established were applied to a reciprocating compressor rig wherein the target was identifying machine states with respect to operational health through comparison of signal outputs for healthy and faulty systems. The technique allowed near perfect fault classification. In addition methods for identifying seperable classes are investigated through profiling techniques, illustrated using Andrew’s Fourier curves.

KW - classifiers

KW - Cluster analysis

KW - Genetic algorithms

KW - reciprocating compressor

KW - Relevance Vector Machine

UR - https://www.scientific.net/AMM.798

U2 - 10.4028/www.scientific.net/AMM.798

DO - 10.4028/www.scientific.net/AMM.798

M3 - Article

VL - 798

SP - 308

EP - 313

JO - Applied Mechanics and Materials

T2 - Applied Mechanics and Materials

JF - Applied Mechanics and Materials

SN - 1662-7482

ER -