Application Study On Fault Diagnosis Of Rotation Mechanical Based On Wavelet And Fractal Method

In accordance with the problems of the fan condition monitoring and fault diagnosis, several kinds of common mechanical faults of the fan is tested in a large fan test-bed and the fault mechanism and its spectrum characteristic are researched. On this basis, using the wavelet theory to de-noise the initial characteristic vibration signal, using wavelet multi-resolution analysis, wavelet packet analysis to analyzing different characteristics of the fault states, make qualitative fault diagnosis;And calculating correlation dimension of the signal of incomes, quantizing characteristics of fault, make quantitative fault diagnosis, realize the discrimination of the fault type, and confirm the sensitive frequency band of the faults.Based on the above analysis, proposing wavelet correlation dimension, wavelet multi-resolution correlation dimension, wavelet packet correlation dimension,wavelet approximate entropy the methods of fault diagnose.Result shows that the above method is superior to the traditional diagnosis method based FFT,which is more visual and capable of specification.It is a kind effective method of monitoring and fault diagnosis of fan.