Research On Fault Diagnosis Method Of Axle Box Bearing Based On Weak Fault Feature Enhancement And Adaptive Variational Modal Decomposition

Axle box bearing plays a key role in bearing and transmission in high-speed train bogie.The research on axle box bearing fault diagnosis is of great practical significance to ensure the safe and stable operation of the train,improve the economy of maintenance and avoid major safety accidents.The vibration signal of axle box bearing is an excellent carrier to reflect its operation state.At the same time,it has the advantages of repeatable monitoring and intuitive and reliable diagnosis results.However,due to the influence of track irregularity excitation and load,its vibration signal is mixed with many interference components,and the relatively weak fault information is easy to be submerged and difficult to extract in the noise.Taking the axle box bearing,the key component of high-speed train bogie,as the object,this paper deeply studies the fault diagnosis method of axle box bearing based on adaptive signal decomposition and signal demodulation analysis under the interference of random noise.The main research contents are as follows:(1)The typical failure forms of axle box bearing are summarized,and the vibration signal characteristics of different types of bearing faults are further analyzed,which provides theoretical support for the follow-up diagnosis method research.(2)In order to effectively enhance the weak fault features in the original vibration signal,a fault feature enhancement method based on autocorrelation noise reduction and local mean decomposition(LMD)is proposed.The enhanced shock index is constructed based on the energy ratio and the kurtosis of the envelope spectrum power spectrum.This method first carries out autocorrelation noise reduction processing on the signal,then carries out LMD decomposition on the processed signal,and further uses the enhanced impact index to screen the effective modes after LMD decomposition to reconstruct the signal and enhance the fault characteristic components in the original signal.The feasibility of the method is verified by experimental data analysis.(3)The parameter setting of variational modal decomposition(VMD)is systematically studied.Through the study of amplitude modulation signal,frequency modulation signal and periodic signal,the different effects brought by the setting of mode number and penalty factor are analyzed.The results show that when the value of mode number is too small,there will be under decomposition,resulting in mode loss or mode aliasing.When the value of mode number is too large,there will be over decomposition,resulting in false modes,and the setting of penalty factor is too small,which will significantly improve the running time of the algorithm,It will also affect the decomposition effect and lead to the phenomenon of mode aliasing.(4)A fault diagnosis method of axle box bearing of high-speed train based on parameter adaptive VMD and fast spectral kurtosis filter is proposed.Taking the enhanced impact index as the fitness function,the sparrow search algorithm is used to adaptively optimize the VMD parameters under different working conditions,and then the VMD decomposed sub signal is filtered by using the fast spectral kurtosis diagram.Finally,the filtered signal is demodulated and analyzed by Hilbert envelope.The superiority and effectiveness of the method are verified by the simulation signal and the data of the axle box bearing test-bed of high-speed train built in this paper.