A New Heavy-Duty Bearing Degradation Evaluation Method with Multi-Domain Features

Area of Science:

• Mechanical Engineering
• Reliability Engineering
• Signal Processing

Background:

• Heavy-duty bearings face complex, variable loads, causing unpredictable degradation patterns.
• Traditional degradation indices struggle to accurately represent bearing health across the entire lifecycle under such conditions.

Purpose of the Study:

• To develop a novel degradation evaluation method for heavy-duty bearings.
• To accurately capture early degradation points and characterize degradation trends.
• To improve the assessment of bearing health under demanding operational conditions.

Main Methods:

• Improved wavelet packet decomposition to extract energy entropy features.
• Construction of a multi-domain feature set combining time and frequency domain data.
• Utilizing a comprehensive evaluation index to form an optimal feature matrix.
• Developing a comprehensive degradation index integrating probability and distance information.

Main Results:

• The proposed method effectively identifies early degradation in bearings.
• It accurately tracks the degradation status of bearings over their lifecycle.
• Validation was performed using two distinct datasets, confirming method efficacy.

Conclusions:

• The new method provides a more accurate assessment of heavy-duty bearing degradation.
• It enables early detection of faults and reliable tracking of bearing health.
• This approach enhances the overall degradation assessment and predictive maintenance capabilities for critical machinery.