In response to the issue of insufficient extraction of effective information due to the effect of environmental noise on weak rolling bearing fault signals in aircraft engines, a method for extracting fault features in rotating machinery main bearings using multi-channel vibration information (MCVI) weighted fusion is proposed. This method first utilizes a weighted fusion model for MCVI to integrate data from multiple vibration sensors into a one-dimensional signal. Subsequently, the fused signal is decomposed using the complete ensemble empirical mode decomposition adaptive noise (CEEMDAN) method. Based on the kurtosis index-correlation coefficient filtering criterion, the impactful components are selected for reconstruction, resulting in a vibration signal rich in bearing fault feature information. Lastly, the weak fault features of bearing faults are identified using the envelope spectrum. Simulation signal identification verification shows that the fault feature energy Q within the envelope spectrum can be increased by 12.4%. The effectiveness of the MCVI weighted fusion method is comprehensively validated based on data from a simulated test bench for intermediate shaft bearings in aero-engines. An analysis of vibration signals from a certain type of aero-engine main bearing demonstrates that the proposed method can effectively extract fault feature information transmitted via complex transmission paths, providing an effective means for processing and diagnosing complex signals related to faults in aero-engine main bearings.

Rolling bearing；Aircraft engine；Multi-channel vibration information；Weighted fusion model；Fault feature extraction；CEEMDAN