Research On Fault Diagnosis In Rotary Machinery Acoustic Emission Signal

Rub-impact between rotor and stator of rotating machines is one of the common faults,It is also a major issue which urgently to be solved.Along with the technology development,the structure of rotating machinery becomes more complex and the clearance between rotor and stator smaller,and thus improves the efficiency of the machines.On the other hand,it may easily lead to the occurrence of rub-impact with huge losses,and which has become the urgent problem waiting to be solved in present.It is necessary to study this fault feature,research new accurate approaches location the rub-impact sources,and explore steady diagnosis methods.All of those have import practical significance to protect machine safe operation and develop the technology of rub-impact fault diagnosis.The theoretical analysis and experimental study is on the basis of acoustic emission and combine with fractal theory,wavelet transform and fourier transform in this paper.The main work and research results are as follows.1.Rub-impact is a common fault in rotating machinery operation.It will be of great significant to warn the rub-impact fault for protecting the safety of unit.In this work,a wavelet fractal method is proposed to identify the rub-impact.The signal of acoustic emission rub-impact can be decomposed into different frequency bands by using wavelet transform and then calculate the variance of each frequency band individually;finally the fractal dimension is acquired.The acoustic emission rub-impact signal is characterized by conditions of no rub-impact,slight rub-impact and heavy rub-impact.Analysis the acoustic emission signal under conditions of no rub-impact,slight rub-impact and heavy rub-impact,calculate the wavelet fractal dimension of the acoustic emission rub-impact signal,and then distinguish the condition of rub-impact based on the change of the fractal dimension.By comparing this method with G-P algorithm,it can be concluded that the fractal method can effectively distinguish the happening of the rub-impact2.Acoustic emission signals are highly susceptible to noise interference in rotating machinery fault diagnosis.The empirical mode decomposition(EMD)associates with mode mixing,this paper achieved a method that de-noising and the fault diagnosis of the rotating machinery AE signal based on empirical wavelet transform.This method takes the advantages of the EMD and wavelet transform,classifying the Fourier spectrum by its adaptive property,constructing the wavelet filter bank to extract the different intrinsic mode components of acoustic emission signal,which can eliminate the mode mixing phenomenon.Then the Hilbert transform was carried on the component of the acoustic emission signal so as to realize the de-noising and fault diagnosis.Adopting this method to de-noising the simulations signal that has been added noise,at the same condition,compared with the result of global threshold value de-noising,default threshold value de-noising,tackle high frequency coefficient de-noising based on dB4 and EMD de-noising.Applying this method in the practical AE rubbing signal.Results showed that:Intrinsic modes of the signal can be decomposed effectively through EWT method,the decomposed mode is less and there is no mode that is difficult to explain.Furthermore,de-noising effect is superior to other methods and has great advantage in AE signal fault diagnosis3.Fourier Decomposition Method is an adaptive signal decomposition method based on Fourier transform.This paper proposed the rotating machinery fault diagnosis method based on FDM.First,searching for the minimum number of Fourier Intrinsic Band Functions(AFIBFs)adaptively in the Fourier frequency domain.Then the non-stationary and nonlinear signal is decomposed to a number of Fourier intrinsic band functions and a residual component.Obtaining the time-frequency energy spectrum by analytic FIBFs.Subsequently,diagnosing the fault according to the change of the time-frequency energy spectrum.The object of study is rotor rub-impact acoustic emission signal;the effectiveness of the method is illustrated by simulation and experiment.Comparison of the results by FDM with the fault analysis based on Empirical Mode Decomposition(EMD).Results show that the non-stationary and nonlinear signal can be effectively divided by Fourier Decomposition Method.This method can diagnose the fault and the time of the fault occur validity.It has high time-frequency distribution and avoids the phenomenon of mode mixing and end effect,providing a new method for rotating machinery fault diagnosis.