Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.493933
Title: Automatic bearing fault diagnostics using wavelet analysis and an artificial neural network
Author: Abdul-Raheem, Khalid Fatihi
ISNI:       0000 0001 3389 8441
Awarding Body: Glasgow Caledonian University
Current Institution: Glasgow Caledonian University
Date of Award: 2009
Availability of Full Text:
Access from EThOS:
Abstract:
Machinery failure diagnosis is an important component of the Condition Based Maintenance (CBM) activities for most engineering systems. Rolling element bearings are the most common cause of rotating machinery failure. The existence of the amplitude modulation and noises in the faulty bearing vibration signal present challenges to effective fault detection method. The wavelet transform has been widely used in signal de-noising due to its extraordinary time-frequency representation capability. A new technique for an automated detection and diagnosis of rolling bearing conditions is presented in this thesis. The time-domain vibration signals of rolling bearings with different fault condition are pre-processed using Impulse and Laplace wavelet transforms for rolling bearing fault detection and feature extraction, respectively. The wavelet denoising and the wavelet envelope power spectrums are used for bearing fault detection and diagnosis. Furthermore, the extracted features for the wavelet transform coefficients in time and frequency domain are applied as input vectors to Artificial Neural Networks (ANN) for rolling bearing fault classification. The Impulse and Laplace Wavelets shape and the ANN classifier parameters are optimized using a genetic algorithm (GA). To reduce the computation cost, decrease the size, and enhance the reliability of the ANN, only the predominant wavelet transform scales are selected for feature extraction. The results for both real and simulated bearing vibration data show the effectiveness of the proposed technique for bearing condition identification and classification with very high success rate using minimum input features.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.493933  DOI: Not available
Share: