Fault Diagnosis for Rolling Bearings Using Optimized Variational Mode Decomposition and Resonance Demodulation

Area of Science:

• Mechanical Engineering
• Signal Processing
• Reliability Engineering

Background:

• Locomotive rolling bearing fault diagnosis faces challenges with low accuracy and difficulty in extracting fault signal features.
• Existing methods often struggle to effectively identify subtle fault indicators in complex vibration data.

Purpose of the Study:

• To propose a novel fault diagnosis method, GNVRFD, for locomotive rolling bearings.
• To enhance the accuracy and effectiveness of fault diagnosis by integrating optimized VMD and resonance demodulation.

Main Methods:

• Developed a parameter optimization algorithm combining genetic algorithm and nonlinear programming to adaptively tune Variational Mode Decomposition (VMD) parameters.
• Utilized optimized VMD to decompose vibration signals into intrinsic mode functions (IMFs) and selected sensitive IMFs based on kurtosis values.
• Applied resonance demodulation technology to the reconstructed signal to obtain the envelope spectrum for fault frequency identification.

Main Results:

• The proposed GNVRFD method successfully decomposed vibration signals and identified key fault frequencies.
• Experimental validation using actual rolling bearing data demonstrated the effectiveness of the GNVRFD method.
• Comparative analysis showed that GNVRFD significantly outperforms other fault diagnosis methods in accuracy and efficiency.

Conclusions:

• The optimized VMD combined with resonance demodulation offers a robust solution for locomotive rolling bearing fault diagnosis.
• The GNVRFD method provides a more accurate and effective approach for detecting bearing faults, crucial for operational safety.
• This study validates the practical applicability and superior performance of the proposed GNVRFD technique.