An Optimal Parameter Selection Method for MOMEDA Based on EHNR and Its Spectral Entropy

Area of Science:

• Mechanical Engineering
• Signal Processing
• Condition Monitoring

Background:

• Rolling bearings are critical in industrial production but prone to failures under complex conditions.
• Accurate health assessment of rolling bearings is essential for maintaining mechanical system operation.
• Existing Multipoint Optimal Minimum Entropy Deconvolution Adjusted (MOMEDA) methods for fault diagnosis have limitations in fault period selection and denoising accuracy.

Purpose of the Study:

• To address the limitations of traditional MOMEDA in rolling bearing fault diagnosis.
• To propose an improved MOMEDA (IMOMEDA) method for more accurate fault feature extraction.
• To enhance the reliability and performance of rolling bearing fault diagnosis systems.

Main Methods:

• Utilizing the envelope harmonic-to-noise ratio (EHNR) spectrum for fault period estimation.
• Implementing an improved grid search method with EHNR spectral entropy to determine optimal filter length.
• Combining the improved MOMEDA (IMOMEDA) with the Teager-Kaiser energy operator (TKEO) for feature extraction.

Main Results:

• The proposed IMOMEDA method effectively estimates the fault period without prior knowledge.
• Optimal filter length selection is achieved using EHNR spectral entropy, improving signal denoising.
• The IMOMEDA-TKEO approach demonstrates superior effectiveness and generalization in rolling bearing fault diagnosis across multiple datasets.

Conclusions:

• The developed IMOMEDA method overcomes key limitations of traditional MOMEDA for rolling bearing fault diagnosis.
• The integration of EHNR spectrum and improved grid search enhances diagnostic accuracy.
• The proposed method offers a robust and reliable solution for rolling bearing condition monitoring.