Fault detection in rotating machinery using kernel-based probability density estimation

M. J. Desforges, P. J. Jacob, A. D. Ball

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

In this paper, a method is presented which allows abnormal or unexpected operating conditions to be identified from measured response data. Potential applications of such a technique cover a wide range of engineering situations where a definite, early warning of an abnormal state is essential, but where classification of the particular abnormality is of lesser importance. In the technique described, unexpected operating conditions are identified by the presence of measured data which are significantly different from those known to correspond to normal operating conditions or responses. The proposed approach to the detection of such novel data is based upon probability density function (PDF) estimation using a kernel method. The theory behind this approach is presented for the multivariate case, and the need for data pre-processing in practical applications of PDF estimation is highlighted. The kernel method is demonstrated on simulated data sets. Finally, vibration response data from an electric motor with three levels of phase imbalance are used to illustrate the application of the PDF estimation method to the detection of abnormal conditions. The proposed method yields a clear indication of the existence of a fault in the AC motor, with no prior knowledge of the particular fault state.

Original languageEnglish
Pages (from-to)1411-1426
Number of pages16
JournalInternational Journal of Systems Science
Volume31
Issue number11
DOIs
Publication statusPublished - 1 Jan 2000
Externally publishedYes

Fingerprint

Rotating Machinery
Rotating machinery
Density Estimation
Fault Detection
Probability Density
Fault detection
Probability density function
Function Estimation
kernel
Kernel Methods
AC motors
Electric motors
Fault
Electric Motors
Data Preprocessing
Early Warning
Prior Knowledge
Processing
Vibration
Cover

Cite this

@article{faeec1f00f8b406084d8e6e84503cd2d,
title = "Fault detection in rotating machinery using kernel-based probability density estimation",
abstract = "In this paper, a method is presented which allows abnormal or unexpected operating conditions to be identified from measured response data. Potential applications of such a technique cover a wide range of engineering situations where a definite, early warning of an abnormal state is essential, but where classification of the particular abnormality is of lesser importance. In the technique described, unexpected operating conditions are identified by the presence of measured data which are significantly different from those known to correspond to normal operating conditions or responses. The proposed approach to the detection of such novel data is based upon probability density function (PDF) estimation using a kernel method. The theory behind this approach is presented for the multivariate case, and the need for data pre-processing in practical applications of PDF estimation is highlighted. The kernel method is demonstrated on simulated data sets. Finally, vibration response data from an electric motor with three levels of phase imbalance are used to illustrate the application of the PDF estimation method to the detection of abnormal conditions. The proposed method yields a clear indication of the existence of a fault in the AC motor, with no prior knowledge of the particular fault state.",
author = "Desforges, {M. J.} and Jacob, {P. J.} and Ball, {A. D.}",
year = "2000",
month = "1",
day = "1",
doi = "10.1080/00207720050197794",
language = "English",
volume = "31",
pages = "1411--1426",
journal = "International Journal of Systems Science",
issn = "0020-7721",
publisher = "Taylor and Francis Ltd.",
number = "11",

}

Fault detection in rotating machinery using kernel-based probability density estimation. / Desforges, M. J.; Jacob, P. J.; Ball, A. D.

In: International Journal of Systems Science, Vol. 31, No. 11, 01.01.2000, p. 1411-1426.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Fault detection in rotating machinery using kernel-based probability density estimation

AU - Desforges, M. J.

AU - Jacob, P. J.

AU - Ball, A. D.

PY - 2000/1/1

Y1 - 2000/1/1

N2 - In this paper, a method is presented which allows abnormal or unexpected operating conditions to be identified from measured response data. Potential applications of such a technique cover a wide range of engineering situations where a definite, early warning of an abnormal state is essential, but where classification of the particular abnormality is of lesser importance. In the technique described, unexpected operating conditions are identified by the presence of measured data which are significantly different from those known to correspond to normal operating conditions or responses. The proposed approach to the detection of such novel data is based upon probability density function (PDF) estimation using a kernel method. The theory behind this approach is presented for the multivariate case, and the need for data pre-processing in practical applications of PDF estimation is highlighted. The kernel method is demonstrated on simulated data sets. Finally, vibration response data from an electric motor with three levels of phase imbalance are used to illustrate the application of the PDF estimation method to the detection of abnormal conditions. The proposed method yields a clear indication of the existence of a fault in the AC motor, with no prior knowledge of the particular fault state.

AB - In this paper, a method is presented which allows abnormal or unexpected operating conditions to be identified from measured response data. Potential applications of such a technique cover a wide range of engineering situations where a definite, early warning of an abnormal state is essential, but where classification of the particular abnormality is of lesser importance. In the technique described, unexpected operating conditions are identified by the presence of measured data which are significantly different from those known to correspond to normal operating conditions or responses. The proposed approach to the detection of such novel data is based upon probability density function (PDF) estimation using a kernel method. The theory behind this approach is presented for the multivariate case, and the need for data pre-processing in practical applications of PDF estimation is highlighted. The kernel method is demonstrated on simulated data sets. Finally, vibration response data from an electric motor with three levels of phase imbalance are used to illustrate the application of the PDF estimation method to the detection of abnormal conditions. The proposed method yields a clear indication of the existence of a fault in the AC motor, with no prior knowledge of the particular fault state.

UR - http://www.scopus.com/inward/record.url?scp=0034334821&partnerID=8YFLogxK

U2 - 10.1080/00207720050197794

DO - 10.1080/00207720050197794

M3 - Article

VL - 31

SP - 1411

EP - 1426

JO - International Journal of Systems Science

JF - International Journal of Systems Science

SN - 0020-7721

IS - 11

ER -