Hierarchical Amplitude-Aware Permutation Entropy-Based Fault Feature Extraction Method for Rolling Bearings

Area of Science:

• Mechanical Engineering
• Signal Processing
• Reliability Engineering

Background:

• Rolling bearings are critical components in machinery.
• Early fault detection is essential for preventing catastrophic failures and reducing maintenance costs.
• Traditional methods struggle with complex vibration signals from degrading bearings.

Purpose of the Study:

• To propose an enhanced method, hierarchical amplitude-aware permutation entropy (HAAPE), for detecting incipient faults in rolling bearings.
• To improve the identification of fault characteristics in complex time series data.
• To evaluate the performance degradation of bearings throughout their lifecycle.

Main Methods:

• Combining hierarchical analysis with amplitude-aware permutation entropy (AAPE) to extract multilevel fault information from vibration signals.
• Applying HAAPE to analyze both low-frequency and high-frequency components of abnormal bearing vibrations.
• Developing a fault feature selection strategy using HAAPE after LCM-SVD processing of fault signals.

Main Results:

• HAAPE demonstrates high sensitivity to early-stage rolling bearing failures.
• The method is effective in evaluating bearing performance degradation during the run-to-failure cycle.
• Comparative analysis shows HAAPE outperforms other entropy-based methods in fault feature extraction accuracy.

Conclusions:

• HAAPE offers a superior approach for early fault detection and characteristic identification in rolling bearings.
• The method provides a reliable tool for assessing bearing health and predicting failures.
• HAAPE enhances the accuracy and effectiveness of condition monitoring systems for rotating machinery.