A Novel Demodulation Analysis Technique for Bearing Fault Diagnosis via Energy Separation and Local Low-Rank Matrix...

Area of Science:

• Mechanical Engineering
• Signal Processing

Background:

• Bearing fault diagnosis is critical for mechanical equipment maintenance.
• Fault vibration signals are often modulated due to bearing characteristics.
• Existing demodulation techniques struggle with noise and multi-component signals.

Purpose of the Study:

• To develop a novel demodulation analysis technique for bearing fault diagnosis.
• To address limitations of existing methods in handling noise and signal complexity.
• To improve the accuracy and reliability of bearing fault detection.

Main Methods:

• Utilizing an energy separation algorithm based on the Teager energy operator for amplitude envelope and instantaneous frequency calculation.
• Introducing a new signal decomposition method based on local low-rank matrix approximation (LLORMA).
• Decomposing signals into single components and simultaneously eliminating noise.

Main Results:

• The proposed technique successfully decomposes complex signals and eliminates noise.
• Identified single-component signals exhibit high-frequency features indicative of bearing faults.
• Demonstrated excellent diagnostic performance on simulated and experimental bearing fault signals.

Conclusions:

• The novel demodulation analysis technique shows excellent diagnostic performance for bearing fault signals.
• LLORMA-based signal decomposition effectively handles noise and complex modulated signals.
• The method enhances mechanical equipment maintenance through reliable fault detection.