【摘要】 针对滚动轴承故障信号在低信噪比情况下，难以提取故障频率问题，提出一种将小波包变换，参数优化变分模态分解（Variational Mode Decomposition,VMD）与最大相关峭度解卷积（Maximum Correlated Kurtosis Deconvolution,MCKD）相结合的轴承故障频率提取方法。采用小波包变换的方法，实现对故障信号的分解和重构，计算重构后各个频段信号能量所占的比例，选择能量比例最高的频段信号利用经蜣螂搜索算法（Dung Beetle Optimizer,DBO）优化，对信号进行自适应分解，构建加权峭度指标以筛选最优模态分量；然后对最优模态分量利用DBO优化后的MCKD进行增强；最后，采用包络解调的方法对增强后的信号进行处理，并对其分析，诊断轴承故障频率。实验验证了所提出的方法实现在强噪声环境下滚动轴承故障频率检测。更多还原

【Abstract】 A bearing fault frequency extraction method is proposed, which combines wavelet packet transform, parameter optimized variational mode decomposition(VMD), and maximum related kurtosis deconvolution(MCKD) to address the difficulty in extracting fault frequencies from rolling bearing fault signals under low signal-to-noise ratios. Using the method of wavelet packet transform, the fault signal is decomposed and reconstructed. The proportion of energy in each frequency band signal after reconstruction is calculated. The frequency band signal with the highest energy proportion is selected and optimized using the dung beetle optimizer(DBO) algorithm. The signal is adaptively decomposed and a weighted kurtosis index is constructed to screen the optimal modal components. Then, the optimal modal components are enhanced using the MCKD optimized by DBO. Finally, the enhanced signal is processed using envelope demodulation method and analyzed to diagnose the frequency of bearing faults. The experiment verified that the method proposed in this article achieves fault frequency detection of rolling bearings in strong noise environments.更多还原