Research On Fault Diagnosis Technology Of Rolling Bearing

With the rapid development of science and technology, modern mechanical equipment is more and more intelligent and complex. Rolling element bearing is an important part of the core machinery and equipment. Rolling bearing complex structure, operating conditions and changeable, some even need to work in high temperature and high pressure and other adverse environment for a long time. Its running status is healthy or not will directly determine whether the normal operation of machinery and equipment. Effective fault diagnosis for complex mechanical equipment to rolling bearing, become an urgent problem to be solved in practical engineering application. This paper used in recent years in speech recognition is widely used in hidden Markov model(Hidden Markov Model, HMM), combined with wavelet analysis feature extraction technology to complete the classification of the operating state of the rolling bearing, so as to realize the fault diagnosis of rolling bearing.Accordingly, designed a fault diagnosis scheme and test platform of rolling bearing, the rolling bearing fault mechanism analysis and the fault diagnosis scheme is completed. Then, through the study of in-depth study of HMM, HMM has a wide development and application in speech recognition.The HMM model does not need the accurate mathematical model of the object to be measured, using a small amount of observation sequences, can quickly establish the fault diagnosis model. The HMM model does not need the accurate mathematical model of the object to be measured, using a small amount of observation sequences, can quickly establish the fault diagnosis model. For the rolling bearing vibration non-stationary signal, it is few highlight the special advantage. Build a fault diagnosis model of rolling bearing based on HMM algorithm, the rolling bearing fault diagnosis, the overall completion of the fault diagnosis of rolling bearing.Secondly, on the research of wavelet analysis feature information extraction technology, expounds the feasibility analysis of the vibration signal of rolling bearing wavelet, and the feature extraction after the characteristic signal on vector quantization. The vibration acceleration signal in engineering as an example, verify the application of wavelet analysis in the feature extraction.Finally, combined with the rolling bearing fault diagnosis technology feature extraction technology and HMM wavelet analysis, the test platform designed in this paper, through the fault of rolling bearings of different fault failure simulation test, data acquisition of vibration acceleration data, the vibration data of different failure modes of the HMM models are established for different, the test data is input to the HMM all the prices of different models, HMM model calculated the likelihood values, the maximum value corresponding state model, is the result of the fault diagnosis data to be detected. In order to complete the identification and classification of fault state of ball bearing, test results showed that the rolling bearing fault diagnosis, the model has a high diagnostic rate of recognition design and good feasibility.