Fault Diagnosis of Fan Bearing Based on CEEMDAN-VMD Fusion Feature and SO-SVM
Since fan bearings are prone to failure and vibration signal analysis is extremely effective for fault diagnosis,this paper proposed a Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEMDAN)and Variational Modal Decomposition(VMD)combined signal processing method.Firstly,CEEMDAN was used to decompose the collected vibration signals into several Intrinsic Mode Function(IMF)components,and the energy-weighted composite kurtosis index was used to screen IMF components with obvious fault characteristics,and the signal was reconstructed.After that,the new signals were decom-posed using VMD,and the energy ratio of each IMF after VMD decomposition was combined with the optimal IMF component screened by the composite index of envelope entropy and envelope spectrum kurtosis to construct energy entropy,sample entropy and approximate entropy for feature fusion.Finally,the fusion fea-ture matrix was input into the Snake optimization algorithm(SO)optimization support vector machine(SVM)for recognition and classification,and multi-fault pattern recognition was realized.The simulation results show that the diagnostic accuracy of this method is 98%for the detection of ten kinds of bearing dete-rioration states.It provides a new way of fault diagnosis for fan bearing.
万方数据
维普
