| Rolling bearings is one of the indispensable power transmission devices in rotating machinery.Once a failure occurs,the operation of equipment cannot be guaranteed,resulting in economic losses and fatalities.Therefore,the research on condition monitoring and fault diagnosis of mechanical equipment and rolling bearings has important theoretical significance and practical value.Adaptive variational mode extraction(AVME)is a newly proposed signal decomposition method.It adaptively sets the center frequency parameters of multi-component modes through signal length and bandwidth,and then adaptively decomposes a complex vibration signal into multiple single component modes by use of cyclic constraints,and finally the signal decomposition problem was transformed into multi-mode optimization problem.With the funding of the National Natural Science Foundation of China,this paper studied deeply on AVME theory.Moreover,corresponding solutions have been proposed to solve the theoretical problems existing in AVME.Through the analysis of simulation signals and actual data,the results demonstrated that the proposed method is effective for condition monitoring and fault diagnosis of mechanical equipment and rolling bearings.Meanwhile,it superior to other existing methods.The main research contents of this thesis are as follows:(1)Aiming at the problem that VME can only extract one component according to a central frequency and can not realize the adaptive decomposition of multi-component signals,a new adaptive variational mode extraction method is proposed.On this basis,by fusing kurtosis,correlation coefficient and orthogonality,a method for selecting the optimal frequency modulation band of rolling bearing based on adaptive variational mode extraction and fusion index is proposed,which is verified by bearing simulated fault signal and measured data The results show that the proposed method has advantages in signal decomposition and fault demodulation accuracy.(2)A rolling bearing fault diagnosis method based on AVME,multi-scale permutation entropy and Cuckoo optimized vector machine is proposed to realize the intelligent classification of bearing fault state.Through the analysis of experimental data,it is compared with the fault diagnosis methods based on Empirical mode decomposition,Ensemble empirical mode decomposition and Variational mode decomposition.The results reveal that the proposed method can effectively classify different bearing fault states and is better than the compared methods.(3)Aiming at the problem that the signal fault characteristics of variable speed rolling bearing vary with speed,a fault diagnosis method of variable speed rolling bearing based on AVME and order spectrum is proposed.In this method,the pulse component of vibration signal is extracted by AVME,and the most abundant component containing sensitive fault information of rolling bearing is selected by envelope entropy.Then,the non-stationary signal is transformed into stationary angle domain signal by order analysis.Finally,the fault information is obtained by envelope order analysis.The analysis results of measured variable speed rolling bearing signals verify the effectiveness of the proposed method.(4)A multivariate adaptive variational mode extraction method is proposed to be suitable for multi-channel signal decomposition.Based on this,a new rolling bearing diagnosis method based on Multivariate Adaptive Variational Mode Extraction is proposed.Aiming at the problem of setting the penalty parameters in the adaptive variational mode extraction method,an improved method based on empirical formula is proposed.The central frequency similarity error threshold is used to ensure the identity,and the frequency band consistency of the adaptive decomposition components of multi-channel signals is realized.Through the comparative analysis of simulation and measured signals,the results show that the proposed method is effective in multivariate signal decomposition and consistent frequency band of fault demodulation,and is better than the compared methods.To sum up,this paper mainly studies the algorithm and application of the adaptive variational mode extraction(AVME)method.The proposed method is adopted to analyze the simulated and measured data,the results verify the effectiveness of this method in rolling bearing fault intelligent classification and variable speed rolling fault diagnosis.It provides a new solution for rolling bearing fault monitoring and diagnosis. |
