Based On Cepstral Analysis Of Rolling Bearing Fault Diagnosis

Cepstrum analysis is a fault diagnosis method based on the Fourier transform. This method has a strong ability of analyzing in the detection of periodic signal, in which the Fourier transform is used to switch the original periodic signal into the discrete components type harmonic signals. Moreover, cepstrum prevents the cycle of vibration energy of the weaker elements being ignored by using Logarithmic transformation.Fatigue spalling failure of bearings is one of the main faults in the chemical rotating machinery. When the bearing fatigue spalling failure occurs somewhere within, or under the action of axe rotation period, the failure of the vibration characteristics of the signal presents periodically and can be detected by cepstrum analysis. However, when the failure's harmonic frequency is similar to other structures'fault characteristics, where the misjudgment is alike.In this paper, cepstrum analysis was used in the bearing fault diagnosis. At first, based on the detailed theoretical research, the advantages and disadvantages of the cepstrum analysis in bearing fault diagnosis was explained. Finally, the optimization of cepstrum analysis was put forward to solve the interference problem. The main research and achievements are as follows:(1) Method research. The cepstrum analysis is accomplished by adopting the Fourier Transform twice. The first transform is to detect the periodic signal, and make the signal to present as Power Spectrum which contains a number of Side-band signals. The fault characteristics have been included in the Side-band signals. For protecting the signals from being ignored, the Logarithmic Transformation has been used, and finally the Fourier Transform has been used again to change it into quefrency. The frequency of original signal has been showed by the reciprocal of quefrency, if the frequency is similar to the structures'fault characteristics, the fault features could be obtained.(2)Object research. The fatigue spalling of bearing signal presents as periodic signal Moreover, the period can be calculated with the working parameters. Thus it can be found that the diagnose of the fatigue spalling by cepstrum analysis is applicable.(3) Experimental research. In order to verifying the effectiveness of the cepstrum analysis, the running bearing's vibration signal should be collected. In this paper the type bearing NU205 was chosen as the diagnostic target,that its outer ring, ball bearings, and inner ring are the places that the fatigue spalling occurs, and QPZZ-11 rotating machinery fault simulation integrated platform has been used as the bearing operating platform, the vibration signal was acquired by PXI-1042Q high-performance test systems. During the acquisition process, sensors and sampling frequency is two very important factors, sensor parameters and its placement will affect the signal acquisition of interference and the error, but if the sampling frequency is too low, the signal distortion and frequency mixing phenomena will probably appear which leads to diagnostic results deviation.(4) Failure analysis. Cepstrum analysis was used in the experimental data analysis. At first, the outer ring of bearing fault signals analysis process was compared with the normal bearing signal to expatiate the cepstrum analysis in actual operation process. In the cepstrum analysis of the bearing outer ring, the fault feature quefrency occurred, which could be converted into the frequency to determine the fault of the outer ring. In the fault analysis of the inner ring and ball bearing, the cepstrum came with the harmonic and noise signals that would cause the misjudgment of the fault location.(5) Method of optimization. Combined with the characteristics of spectral analysis, the cepstrum optimization is the third Fourier transform optimization method, which effectively eliminates noise and harmonics signals into a single line. In addition, the failure analysis of the ball bearing and the inner ring had proved ulteriorly that the harmonic fault can be transformed into more obvious information by this method, but also comes with reduced noise interference in the cepstrum, in this way the precise positioning of bearing fatigue spalling failure can be achieved.