An Improved Iterative Filter Decomposition Method And Its Application In Rolling Bearing Fault Diagnosis

Research on fault diagnosis technology of rolling bearings is of great significance for ensuring the safety of mechanical and electrical equipment and reducing economic losses.Most vibration signals of rolling bearings are non-stationary signals.Time-frequency analysis method can extract local information of vibration signals in both frequency and time domains,which is suitable for fault diagnosis of rolling bearings.However,Winger-Ville distribution,short-time Fourier transform,wavelet transform and empirical mode decomposition all have shortcomings in adaptive signal methods,so it is necessary to study new rolling bearing fault methods.In this paper,the theory of iterative filtering decomposition method and its application in fault identification of rolling bearings are explored.Its main contents are as follows:For the rolling bearing method,on the basis of discussing the research status of its common diagnostic methods,the existing problems are found out and the preparation for further research is made.By exploring the theory of iterative filtering,the decomposition effect of iterative filtering is simulated and analyzed,and an iterative filtering method based on mirror continuation is proposed,which solves the end-effect defect of iterative filtering method.The decomposition effect of iterative filtering method and EMD is compared by simulation signal decomposition,and it is found that iterative filtering method has more advantages in overcoming end-effect defect and mode aliasing.And it has better adaptability to the decomposition of noise.Aiming at the difficulty of artificial selection of filtering frequency band in resonance demodulation method,a rolling bearing fault diagnosis method based on iterative filtering decomposition and power spectrum is proposed,which is verified by experimental signals and real engineering signals.The filtering characteristics of bearing signal decomposition using iterative filtering decomposition are analyzed.The method is simple and has application value.At the same time,the method is compared with the power spectrum diagnosis method based on EMD,and the results show that the former is better than the EMD and power spectrum based bearing identification method in extracting bearing damage characteristics;but the bearing fault diagnosis method based on iterative filter decomposition has achieved good results in extracting the inner and outer ring fault of rolling bearing,but in extracting bearing rolling element fault features.Invalid.Aiming at the problem that noise signal affects the effect of iterative filtering decomposition,a bearing diagnosis method based on iterative filtering and optimal minimum entropy deconvolution is proposed and introduced into bearing fault diagnosis.The bearing signal denoising process using optimal minimum entropy deconvolution can improve the diagnostic effect of the iterative filtering method.For the fault bearing signal,the iterative filtering decomposition is used first,and then the bearing signal is decomposed.The optimal minimum entropy denoising of the key components can accurately match the bearing fault characteristics,and the extraction effect is good.This method has good effect in extracting rolling element faults and is superior to envelope spectrum analysis.To solve the problem of low signal-to-noise ratio of bearings,a rolling bearing fault diagnosis method based on iterative filtering and envelope spectrum is proposed.First,the bearing damage signal is decomposed,and then the key eigenvalue components are analyzed by envelope spectrum,which can effectively extract the bearing fault characteristics.This method is simple and practical.The comparison shows that this method is superior to the single envelope spectrum method.Finally,the validity of the method is verified by the fault data of the inner and outer circles of West Chu School in the United States.