We present a feasibility analysis for the development of an online ball bearing fault detection and identification method which can effectively classify various fault stages related to the contact in the coated ball bearings using vibration measurements. To detect ball bearing faulty stages, we have developed new degree of randomness (DoR) analysis methods using Shannon entropy and random covariance matrix norm theory. To classify the fault stages, we have further developed a set of stochastic models using Gaussian Mixture Hidden Markov Model (GM-HMM) theory. Test results have shown that our algorithms can predict bearing failures without using actual failure data.
Online coated ball bearing health monitoring using degree of randomness and Hidden Markov Model
2009-03-01
1026731 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Online adaptive hidden Markov model for multi-tracker fusion
British Library Online Contents | 2016
|Online adaptive hidden Markov model for multi-tracker fusion
British Library Online Contents | 2016
|Online adaptive hidden Markov model for multi-tracker fusion
British Library Online Contents | 2016
|Online Identification of Hidden Semi-Markov Models
British Library Conference Proceedings | 2003
|