参数自适应变分模态分解的滚动轴承故障诊断OA

To effectively identify rolling bearing faults and address the adaptive optimization of Variational Mode Decomposition(VMD)parameters,the author proposed using Global Search Whale Optimization Algorithm(GS-WOA)to optimize VMD parameters,combined with Weighted Ensemble Kurtosis(WEK)to select the optimal Intrinsic Mode Function(IMF),thereby achieving effective diagnosis of rolling bearing faults.Firstly,the amplitude of each mode was analyzed using the Hilbert marginal spectrum to achieve the optimal adjustment of the VMD parameter range.Secondly,the minimum fuzzy entropy was adopted as the fitness function of GS-WOA for VMD decomposition,the optimal intrinsic mode component was selected by maximizing WEK,and envelope spectrum analysis was performed on this component to accurately identify the type of rolling bearing fault.Finally,the bearing simulation signal and the actual measured inner ring fault data were collected for experimental verification to validate the effectiveness and practicality of the proposed method.The results show that compared with the sequential variational mode decomposition and the empirical mode decomposition method,the envelope spectrum obtained by this method is clearer,and the fault frequency is more obvious.The fault frequency of the inner ring was 162.19 Hz.This method provides an adaptive analysis approach for bearing fault diagnosis that does not require manual preset parameters,has high accuracy and strong anti-noise ability.