Bearing Fault Diagnosis Using Piecewise Aggregate Approximation and Complete Ensemble Empirical Mode Decomposition...

Area of Science:

• Mechanical Engineering
• Signal Processing
• Condition Monitoring

Background:

• Rolling bearing faults are critical in machinery, necessitating accurate diagnostic methods.
• Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) is effective but computationally intensive and memory-demanding.
• Existing CEEMDAN methods struggle with long signal decomposition and efficiency.

Purpose of the Study:

• To develop a more efficient and effective rolling bearing fault diagnosis method.
• To address the limitations of CEEMDAN in terms of computational cost and memory usage.
• To improve the diagnostic accuracy for rolling bearing faults using a novel signal processing technique.

Main Methods:

• A novel rolling bearing fault diagnosis method combining Piecewise Aggregate Approximation (PAA) with CEEMDAN is proposed.
• The method involves generating a vibration envelope via bandpass filtering and demodulation.
• The envelope signal is then compressed using PAA before decomposition with CEEMDAN.

Main Results:

• The proposed PAA-CEEMDAN method demonstrates enhanced diagnostic capabilities for rolling bearing faults.
• The integration of PAA significantly improves the efficiency and reduces the memory requirements of CEEMDAN.
• Verification using test data confirms the superior effectiveness and efficiency compared to standard CEEMDAN.

Conclusions:

• The PAA-CEEMDAN method offers a promising advancement for rolling bearing fault diagnosis.
• This approach effectively overcomes the computational and memory limitations of traditional CEEMDAN.
• The enhanced method provides a more efficient and accurate solution for condition monitoring of rolling bearings.