An improved Autogram and MOMEDA method to detect weak compound fault in rolling bearings

Area of Science:

• Mechanical Engineering
• Signal Processing
• Condition Monitoring

Background:

• Weak compound faults in rolling bearings present challenges due to noise interference and feature coupling.
• Accurate diagnosis is crucial for preventing catastrophic failures and ensuring operational reliability.

Purpose of the Study:

• To develop a novel method for diagnosing weak compound faults in rolling bearings.
• To effectively denoise signals and decouple coupled fault features for improved identification.

Main Methods:

• An improved Autogram, incorporating multi-scale permutation entropy, identifies optimal resonance frequency bands for preliminary denoising.
• Multipoint Optimal Minimum Entropy Deconvolution Adjusted (MOMEDA) is applied to decouple compound fault features.
• Square envelope analysis is used to identify fault characteristic frequencies in the separated signals.

Main Results:

• The proposed method successfully analyzed simulated and experimental datasets of rolling bearing weak compound faults.
• Accurate diagnosis of weak compound faults was achieved.
• The method demonstrated effectiveness and superiority when compared to the parameter-adaptive variational mode decomposition algorithm.

Conclusions:

• The developed method provides an effective solution for diagnosing weak compound faults in rolling bearings.
• The combination of improved Autogram and MOMEDA enhances the ability to separate and identify complex fault signatures.
• This approach offers a significant advancement in bearing condition monitoring and fault diagnosis.