改进辛几何模态分解的滚动轴承故障特征提取

Fault Feature Extraction of Rolling Bearing Based on Improved Symplectic Geometric Modal Decomposition

针对滚动轴承故障冲击信号难以提取的问题,提出了一种改进辛几何模态分解(Improved Symplectic Geometry Modal Decomposition,ISGMD)滚动轴承故障特征提取方法。首先将振动信号进行辛几何模态分解,然后,利用k均值聚类的方法对分解得到的辛几何分量进行聚类,通过包络谱稀疏度指标筛选出故障特征明显的聚类辛几何分量(Cluster Symplectic Geometry Component,CSGC)并进行重构,对重构分量进行包络解调,提取出故障特征。将该方法运用到轴承故障仿真和实验信号,结果表明,这里提出的方法能够有效提取出滚动轴承故障特征。

Aiming at the problem that it is difficult to extract the impact signal of rolling bearing faults,a method based on ISGMD for rolling bearing fault feature extraction is proposed.First,the vibration signal was decomposed by symplectic geomet-ric modal,and then the symplectic geometric components obtained by the decomposition were clustered using the k-means cluster-ing method,and the cluster symplectic geometric components with obvious fault characteristics were screened out through the en-velope spectrum sparsity index CSGC and reconstructed it,and performed envelope demodulation on the reconstructed component to extract the fault characteristics.Applying this method to bearing fault simulation and experimental signals,the results show that the method proposed in this paper can effectively extract the features of rolling bearing faults.