Bearing Fault Diagnosis Method Based on RCMFDE-SPLR and Ocean Predator Algorithm Optimizing Support Vector Machine.

Area of Science:

• Mechanical Engineering
• Data Science
• Signal Processing

Background:

• Accurate extraction of rolling bearing fault characteristics is challenging.
• Existing fault diagnosis methods often suffer from low accuracy.
• The need for robust and efficient fault diagnosis in rotating machinery is critical.

Purpose of the Study:

• To propose an unsupervised method for effective rolling bearing fault diagnosis.
• To enhance the accuracy and stability of fault characteristic extraction.
• To develop a superior fault diagnosis model by combining advanced feature selection and optimization techniques.

Main Methods:

• Utilizing refined composite multiscale fluctuation-based dispersion entropy (RCMFDE) for stable and accurate entropy feature extraction.
• Employing self-paced learning and low-redundant regularization (SPLR) for dimensionality reduction and effective feature selection.
• Implementing a support vector machine (SVM) classifier optimized by the marine predator algorithm (MPA) for fault diagnosis.

Main Results:

• The RCMFDE method demonstrated improved stability and accuracy in bearing characteristic extraction.
• The SPLR technique effectively reduced feature redundancy and enhanced characteristic effectiveness.
• The MPA-optimized SVM model achieved a high bearing fault diagnosis accuracy of 97.67%.

Conclusions:

• The proposed RCMFDE and SPLR combined with MPA-SVM offers a superior approach for rolling bearing fault diagnosis.
• The method effectively addresses challenges in characteristic extraction and diagnosis accuracy.
• This technique shows significant potential for industrial applications requiring reliable machinery health monitoring.