| The flexible thin-walled bearing is a key component of the harmonic reducer,the slight damage of the bearing will not only affect the performance of the harmonic reducer,but also affect the safety and efficiency of the entire mechanical system.Compared with ordinary rolling bearing,flexible thin-walled bearing has special working methods and application occasions,which makes the vibration signal generated by flexible thin-walled bearing during actual operation is more complicated than ordinary rolling bearing.It not only has nonlinear and nonstationary characteristics,but also has both strong and weak fault impact components,and the weak fault impacts will be mixed in the normal impact components generated by the alternate of long and short axes of the cam ellipse,which brings great difficulty to the fault diagnosis of flexible thin-walled bearing.In view of the above problems,this paper takes flexible thinwalled bearing as the research object,takes the kinematics of flexible thin-walled bearing as the research basis,by collecting and analyzing the vibration signals of flexible thin-walled bearing,summarizes a fault diagnosis method of flexible thin-walled bearing based on AFD and EEMD-HHT algorithm.The specific research contents of the paper are as follows:First,analyze the kinematics characteristic of flexible thin-walled bearing,deduce the formula of fault characteristic frequency of flexible thin-walled bearing,introduce the vibration test bench of the flexible thin-walled bearing,and analyze the fault vibration signals of the inner and outer rings of the flexible thin-walled bearing as well as the normal vibration signals collected by this test bench.Secondly,the Adaptive Fourier Decomposition(AFD)algorithm is used to extract the nonlinear low-frequency signals in the fault signals of flexible thin-walled bearing.The results show that the AFD algorithm can effectively extract the low-frequency shock components in the signals and realize the separation of the fault shock and the normal shock.Thirdly,introduce the basic principle of Hilbert-Huang Transform(HHT),and use the EEMD algorithm instead of the EMD algorithm,through the analysis of the inner and outer ring of the fault vibration signals of flexible thin-walled bearing,the results show that the spectral envelope of signal obtained by EEMD-HHT algorithm has a more evident failure frequency spectral line and fewer hybrids.The extraction effect is better than HHT algorithm.Finally,combining the AFD algorithm and EEMD-HHT algorithm,propose a new fault diagnosis method of flexible thin-walled bearing,through the collection of relevant test data and the diagnosis analysis of flexible thin-walled bearing,the characteristic frequency of bearing failure can be accurately and intuitively found in both Hilbert envelope spectrum and Hilbert marginal spectrum of signals,compared with the EEMD-HHT algorithm,the extraction effect is better. |
