Research On Gearbox Fault Diagnosis Based On Variational Mode Decomposition

Gearbox is an important part of mechanical equipment.Its main role in the transmission system is to carry out power transmission and conversion.The safety and reliability of gearbox during operation determine the safety of mechanical equipment.Therefore,it is of great significance to monitor and diagnose the running state of gearbox.Based on the method of variational mode decomposition,the rolling bearing of gearbox is taken as the research object and the faults of rolling bearing are studied.The specific research contents are as follows:(1)The methods of VMD and EMD are compared and analyzed.Compared with EMD,VMD can effectively avoid modal aliasing,which proves that VMD has some advantages in fault diagnosis compared with EMD.The number of modes and penalty factor parameters of VMD method will affect the decomposition effect of VMD on vibration signals.Therefore,the parameters of VMD are optimized based on Genetic Algorithm(GA)and traditional optimization algorithm.Through the analysis of simulation signals,the theoretical basis for the application of VMD parameter optimization method in gearbox fault diagnosis is established.(2)A fault feature extraction method for gearbox bearing with optimal sensor placement is constructed.The mode shapes of gearbox upper box are extracted.The K-medoids algorithm is used to cluster the mode shapes.The effective independent method is applied to initially select the measuring points.Modal Assurance Criterion(MAC)is utilized as the fitness function.Genetic Algorithm(GA)is used to optimize the sensor position.The comprehensive evaluation index is constructed to evaluate the number of sensors so as to obtain the optimal layout scheme of gearbox sensor.The sensors are arranged according to the comprehensive evaluation index.The sensors are arranged in the upper box to collect the vibration signals of gearbox.Each signal is denoised through singular value decomposition(SVD).The denoised signals are fused on the basis of the variance contribution rate.The fusion signal is decomposed through VMD.The VMD parameters are optimize based on the principle of the minimum information entropy of IMF component.The IMF component with the minimum information entropy is selected for Teager energy spectrum analysis to extract the fault features of rolling bearing.The test results on the ZDH-10 gearbox setup show that the fault feature extraction method of gearbox bearing based on optimal sensor placement is effective.(3)A compound fault diagnosis method is constructed to optimize the parameters of MCKD-VMD.Based on the maximum entropy of the filtered signal,the filter length parameters of the MCKD method are optimized,and the composite fault signal is filtered by the MCKD method with optimized parameters to separate the composite fault.The number of modes and the penalty factor of the VMD are optimized with the maximum average envelopekurtosis of the IMF component and the minimum of the average sample entropy respectively.The filtered signal is optimized through the analysis of the composite fault simulation signal and the rolling bearing composite fault test signal from Case Western University.The results show that the method of gearbox composite fault diagnosis based on the optimized parameters of MCKD-VMD is effective.