基于 CP 分解的滚动轴承微弱故障特征提取^∗

To address the limitations that single sensor information acquisition has in characterizing bearing fault and susceptible interference from background noise，a method for extracting weak fault features in rolling bearings based on tensor canonical polyadic（CP）decomposition is proposed in this study. First，based on the cyclostationary characteristics of bearing fault pulse signals under stable operating conditions，the spectral correlation（SC）analysis method is used to transform the multi-channel measurement signals into the SC domain. Subsequently， the multi-channel SC matrices are organized into a tensor indexed by frequency， cyclic frequency，and channel. CP decomposition is then utilized to extract the fault information tensor，and the resulting fault feature tensor is averaged along the channel dimension to obtain an SC matrix that more effectively characterizes the fault features. Finally，a designed filter and the enhanced envelope spectrum are used to further enhance the fault feature SC matrix，the effectiveness of the proposed method is verified through simulations and experiments. The results demonstrate that the proposed method can accurately and effectively extract weak fault features from bearing fault signals under strong background noise interference.