Location:Home > Engineering science >  Control Theory and Control Engineering > A Approach to Fault Feature Extraction of Rolling Bearing Based on Statistical Distribution Model
Details
Name

A Approach to Fault Feature Extraction of Rolling Bearing Based on Statistical Distribution Model

Downloads: []
Author
Tutor: PengTao
School: Hunan University
Course: Control Theory and Control Engineering
Keywords: feature extraction,weibull distrubtion,lognormal distrubtion,wavelet-domain,roll
CLC: TH165.3
Type: Master's thesis
Year:  2011
Facebook Google+ Email Gmail Evernote LinkedIn Twitter Addthis

not access Image Error Other errors

Abstract:
The rolling bearing is one of the most widely used and damageable parts in rotating machinery with its working state directly impacting on the performance of the whole machine, and the fault diagnosis of rolling bearing has a very important practical significance.The fault diagnosis of rolling bearing based on vibration signal analysis is one of the most commonly used and effectively methods.The statistical distribution model parameters are widely used for characterizing the fatigue life and strength of the mechanical products in the reliability engineering. But it has not been paid any attention to that they are used for feature extraction of fault information in the condition monitor and fault diagnosis of mechanism especially the rolling bearing. A novel approach to fault feature extraction based on Weibull distribution parameters is proposed, to mining fully the useful information of state change in the original signal of bearing vibration. After the original signal is modelled as the Weibull distribution, its scale parameter and digital feature-median is extracted as a new feature variable for the state of bearing running respectively. The tests results of fault diagnosis of the rolling bearing verify that this new feature variable is effective.According to the problem that the rolling bearing vibration signal is non-gaussian.A novel approach to fault feature extraction based on lognormal distribution parameters is proposed,its log mean parameter is extracted as a new feature variable for the state of bearing running. To solve the problem of the vibration signals¡¯s non-gaussian.Since the non-stationary of rolling bearing vibration signals can not be fully described by the statistical distribution model parameters,a feature extraction approach based on wavelet-domain lognormal model is proposed. First of all, After adopting wavelet analysis to decompose the nonstationary vibration signal into stationary signals,a non-gaussion distribution¡ªlognormal distribution model of the vibration signal is established,finally, its log mean and log variance are extracted and then constructed as the feature vectors to characterise state of roller bearing running. The tests results of fault diagnosis of the rolling bearing verify that this new feature variable is effective and superiority.
Related Dissertations
Last updated
Sponsored Links
Home |About Us| Contact Us| Feedback| Privacy | copyright | Back to top