Research On The Adaptive Fault Diagnosis Of Bearing Based On Variational Mode Decomposition And Stochastic Resonance

Printing machinery has high operating speed,manufacturing precision and automation.Its running state directly affects the quality of printed products and work efficiency.Therefore,the printing equipment urgently requires advanced and effective fault diagnosis methods.This paper takes the core components of the printing equipment transmission system-bearings,as the research object.Aiming at the problem of the strong background noise,the non-linear and non-stationary vibration feature signal of fault bearing is difficult to extract,the fault diagnosis under small samples and poor correlation.Research on the adaptive fault diagnosis method of printing machine bearing based on variational mode decomposition and stochastic resonance.The concrete research contents are as follows:(1)Research on the adaptive fault diagnosis method of bearing based on variational mode decomposition.It solves the problem that the non-linear and non-stationary vibration feature signal of fault bearing is difficult to extract.By introducing the variational modal decomposition method(VMD).In order to overcome the disadvantage of artificial setting parameters and strong subjectivity in VMD,then an adaptive variational mode decomposition method(AVMD)is proposed.The balance constraints parameters in AVMD mainly affect the decomposition performance.In this paper,VMD parameters are adaptively selected by the power spectrum distribution ratio.The AVMD method is superior to the EMD,EEMD method in terms of orthogonality and energy conservation.It is verified by simulation signal and experiment data.(2)Research on the the adaptive fault diagnosis method of printing machine bearing based on variational mode decomposition and stochastic resonance(AVMD-SR).It solves the problem that the weak signal extraction under strong noise background.By constructing the characteristic coefficients of impulse signal,the SR parameters can be extracted adaptively.The optimized SR and AVMD were combined.It overcomes the limitation of single parameter optimization in SR and neglects the inter-system parameter interaction.Through the simulation experiment data and the offset press working condition data,the results show that not only the fault characteristic frequency is effectively identified,but also the fault amplitude is increased by 4.06 times compared with the the unoptimized method.(3)Research on the small sample fault diagnosis method based on adaptive variational mode decomposition and self-organizing feature mapping network(AVMD-SOFM).It solves the problem that high precision of the printing equipment,complex internal structure,relatively difficult disassembly and less data amount of fault samples.By calculating the permutation entropy of each IMF after AVMD decomposition,putting the entropy into the trained SOFM network to diagnose the fault result.The results show that the accuracy of bearing fault diagnosis based on AVMD-SOFM method is 95%.