Research On Fault Diagnosis Technology Of Gear Box Based On Vibration Signal

As a core component of mechanical facilities,gearboxes have been widely used in the field of machinery.And its performance in the harsh and complex working environment will directly affect the operation of the entire mechanical facilities.If the gearbox has problems during normal operation,it will not only affect normal production,affect economic benefits,and even threaten personal safety.Therefore,it is of great significance to carry out effective safety inspections and regular maintenance on the basis of ensuring that the mechanical facilities can maintain normal operation and have no hidden dangers.The object of this experiment is a laboratory gearbox.Three diagnostic methods were used to diagnose the bearings and gears in the gearbox.In theory,the vibration mechanism of the rolling bearing and the vibration mechanism of the gear are analyzed.The main types of failures involved in the experiment are the inner and outer ring failures of the rolling bearing and the gear tooth failure.The HG8916 equipment is used to collect the characteristic signals of the gearbox under normal working conditions and various fault conditions,and the obtained signal data is stored in a text format.Based on the vibration signal,on the laboratory gearbox dynamics simulation experiment platform,the laboratory gearbox is analyzed in depth,and the MATLAB software system is used to process the various failure signals.In this paper,the bearings in the laboratory gearbox are separately analyzed.And gears use different time-frequency domain analysis methods for fault diagnosis research,which uses complete integrated empirical mode decomposition(CEEMDAN)assisted fast spectral kurtogram algorithm(Fast Kurtogram)and maximum correlation kurtosis deconvolution(MCKD)combined with correlation coefficient enhancement The singular value(SVD)method makes an experimental analysis of the bearing signal,and uses wavelet analysis combined with fast independent component analysis(Fast ICA)to make a fault diagnosis analysis of the gear signal.CEEMDAN-Fast Kurtogram uses a band-pass filter to convert the experimental signal into a digital signal for diagnosis;MCKD-correlation coefficient enhancement SVD is difficult to extract the weak faults of rolling bearings.MCKD is used to enhance the vibration characteristics of the signal,and then the correlation coefficient SVD is used in When decomposing a signal,you can construct a characteristic signal by finding a component that has a large positive correlation with the original signal;wavelet analysis-Fast ICA uses wavelet threshold noise reduction and Fast ICA to perform blind source separation on the noise reduction signal and the original signal to obtain a more characteristic separation signal.