Radial basis function neural network for effective condition monitoring of rolling element bearing

G. S. Vijay, P. Srinivasa Pai, N. S. Sriram, B. K.N. Raj Rao

Research output: Contribution to journalArticle

Abstract

The performance of the Radial Basis Function Neural Network (RBFNN) in the defect classification of a Rolling Element Bearing (REB) has been investigated in this work. The features (input) required for training the RBFNN have been extracted from the non-overlapping segments of the raw and denoised bearing vibration signals. A kurtosis based wavelet denoising method has been used to reduce the noise components in the vibration signals. The Fisher's Criterion (FC) has been used to select a few sensitive features and form a reduced feature set. The centers of the RBF units have been optimized using a modified Fuzzy C-Means (FCM) algorithm, viz., Cluster Dependent Weighted FCM (CDWFCM). The performance of the RBFNN has been compared for four training strategies: two types of feature sets (all features and FC selected features) and two types of RBF centers selection method (centers selected randomly and centers selected using CDWFCM). These strategies have also been tested for the bearing vibration signals provided by the Case Western Reserve University database.

Original languageEnglish
Pages (from-to)21-31
Number of pages11
JournalInternational Journal of COMADEM
Volume18
Issue number3
Publication statusPublished - 01-07-2015

Fingerprint

Bearings (structural)
Condition monitoring
Neural networks
Defects
Radial basis function
Vibration

All Science Journal Classification (ASJC) codes

  • Bioengineering
  • Signal Processing
  • Safety, Risk, Reliability and Quality
  • Strategy and Management
  • Industrial and Manufacturing Engineering
  • Electrical and Electronic Engineering

Cite this

Vijay, G. S. ; Srinivasa Pai, P. ; Sriram, N. S. ; Raj Rao, B. K.N. / Radial basis function neural network for effective condition monitoring of rolling element bearing. In: International Journal of COMADEM. 2015 ; Vol. 18, No. 3. pp. 21-31.
@article{f153303913274f128224b206b7c8c2b0,
title = "Radial basis function neural network for effective condition monitoring of rolling element bearing",
abstract = "The performance of the Radial Basis Function Neural Network (RBFNN) in the defect classification of a Rolling Element Bearing (REB) has been investigated in this work. The features (input) required for training the RBFNN have been extracted from the non-overlapping segments of the raw and denoised bearing vibration signals. A kurtosis based wavelet denoising method has been used to reduce the noise components in the vibration signals. The Fisher's Criterion (FC) has been used to select a few sensitive features and form a reduced feature set. The centers of the RBF units have been optimized using a modified Fuzzy C-Means (FCM) algorithm, viz., Cluster Dependent Weighted FCM (CDWFCM). The performance of the RBFNN has been compared for four training strategies: two types of feature sets (all features and FC selected features) and two types of RBF centers selection method (centers selected randomly and centers selected using CDWFCM). These strategies have also been tested for the bearing vibration signals provided by the Case Western Reserve University database.",
author = "Vijay, {G. S.} and {Srinivasa Pai}, P. and Sriram, {N. S.} and {Raj Rao}, {B. K.N.}",
year = "2015",
month = "7",
day = "1",
language = "English",
volume = "18",
pages = "21--31",
journal = "International Journal of COMADEM",
issn = "1363-7681",
publisher = "COMADEM International",
number = "3",

}

Radial basis function neural network for effective condition monitoring of rolling element bearing. / Vijay, G. S.; Srinivasa Pai, P.; Sriram, N. S.; Raj Rao, B. K.N.

In: International Journal of COMADEM, Vol. 18, No. 3, 01.07.2015, p. 21-31.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Radial basis function neural network for effective condition monitoring of rolling element bearing

AU - Vijay, G. S.

AU - Srinivasa Pai, P.

AU - Sriram, N. S.

AU - Raj Rao, B. K.N.

PY - 2015/7/1

Y1 - 2015/7/1

N2 - The performance of the Radial Basis Function Neural Network (RBFNN) in the defect classification of a Rolling Element Bearing (REB) has been investigated in this work. The features (input) required for training the RBFNN have been extracted from the non-overlapping segments of the raw and denoised bearing vibration signals. A kurtosis based wavelet denoising method has been used to reduce the noise components in the vibration signals. The Fisher's Criterion (FC) has been used to select a few sensitive features and form a reduced feature set. The centers of the RBF units have been optimized using a modified Fuzzy C-Means (FCM) algorithm, viz., Cluster Dependent Weighted FCM (CDWFCM). The performance of the RBFNN has been compared for four training strategies: two types of feature sets (all features and FC selected features) and two types of RBF centers selection method (centers selected randomly and centers selected using CDWFCM). These strategies have also been tested for the bearing vibration signals provided by the Case Western Reserve University database.

AB - The performance of the Radial Basis Function Neural Network (RBFNN) in the defect classification of a Rolling Element Bearing (REB) has been investigated in this work. The features (input) required for training the RBFNN have been extracted from the non-overlapping segments of the raw and denoised bearing vibration signals. A kurtosis based wavelet denoising method has been used to reduce the noise components in the vibration signals. The Fisher's Criterion (FC) has been used to select a few sensitive features and form a reduced feature set. The centers of the RBF units have been optimized using a modified Fuzzy C-Means (FCM) algorithm, viz., Cluster Dependent Weighted FCM (CDWFCM). The performance of the RBFNN has been compared for four training strategies: two types of feature sets (all features and FC selected features) and two types of RBF centers selection method (centers selected randomly and centers selected using CDWFCM). These strategies have also been tested for the bearing vibration signals provided by the Case Western Reserve University database.

UR - http://www.scopus.com/inward/record.url?scp=84959285746&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84959285746&partnerID=8YFLogxK

M3 - Article

VL - 18

SP - 21

EP - 31

JO - International Journal of COMADEM

JF - International Journal of COMADEM

SN - 1363-7681

IS - 3

ER -