The Research Of Fault Diagnosis Of Rolling Bearing Based On Vibration Signal Analysis Method

The rolling bearing is the most commonly used mechanical parts in rotating machinery. The current research on rolling bearing fault diagnosis field has set off a boom, related theories and technologies are developing rapidly, and research results emerge in an endless stream. While the use of vibration signal processing technologyis for rolling bearing fault diagnosis is an ideal method. This paper just was to study the rolling bearing fault diagnosis based on vibration signal processing method, detailed Signal noise reduction, kurtosis index, wavelet theory, wavelet packet energy distribution theory,etc. Finally the two kinds of signal processing method for fault diagnosis were puted forward, the first method was the combination of the kurtosis, wavelet and envelope spectrum analysis method for fault diagnosis of rolling bearing; the another method was the combination of wavelet packet transform, Calculation of frequency band energy and envelope spectrum analysis for fault diagnosis of rolling bearing. In this paper, the contents and conclusions are as follows:1. The fault classification and fault mechanism, vibration of rolling bearings on the vibration causes were pointed out, then from several aspects of the time domain analysis, frequency domain analysis and frequency domain analysis, signal processing method was studied. Found some time domain index especially the kurtosis is more sensitive the early failure of rolling bearing, and can be used to judge the early stage fault processing of signal.2. A comparative study on several time and frequency domain analysis method in the processing of nonstationary signal was conducted, pointed out their Shortcomings, highlighted the wavelet technology is strong efficient in the field of fault diagnosis. aiming at the characteristics of rolling bearing signal, by the combining of wavelet packet decomposition and the energy analysis method to study the signals of different frequency band, quickly sorted out of the energy concentration, representative band signal of containing abundant fault information, reduced the engineering analysis and calculation.3. With rolling bearing test rig data, extracted the bearing vibration signal in the four operation modes. By combining the kurtosis analysis, wavelet decomposition and Hilbert envelope spectrum analysis method for signal noise reduction, fault feature extraction and recognition, finally verified by experiments, this method can effectively identify the types of rolling bearing fault.4. Wavelet analysis can not be more detailed classification of the signal frequency band, similarly to the rolling bearing vibration signal of four different modes of operation, another method of combining wavelet packet decomposition, frequency band energy calculation and Hilbert envelope spectrum analysis were putted forward. This technique can quickly locate rich bands of Containing abundant fault information, can effectively analyze the fault characteristic of periodic impact class. Finally, through the test, this method can strongly extract the features of different fault.Proved through theoretical and experimental verification, wavelet transform related theory, can be applied effectively in fault analysis to identify the rolling bearing, the further research in the vibration signal processing also neends Further study.