A Multivariate Blaschke-Based Mode Decomposition Approach for Gear Fault Diagnosis.

Area of Science:

• Mechanical Engineering
• Signal Processing
• Condition Monitoring

Background:

• Existing multivariate signal decomposition methods lack mechanical insight into gear systems, hindering effective fault feature extraction.
• Accurate fault diagnosis in gear systems is crucial for preventing catastrophic failures and ensuring operational reliability.

Purpose of the Study:

• To propose a novel Multivariate Blaschke-based Mode Decomposition (MBMD) method for enhanced gear system fault diagnosis.
• To address the limitations of current methods by incorporating mechanical characteristics of gear systems into signal decomposition.

Main Methods:

• Modeled multivariate vibration signals as multi-dimensional gear system responses.
• Utilized Stochastic Adaptive Fourier Decomposition (SAFD) for signal representation via Blaschke products, enabling adaptive multi-channel information fusion.
• Introduced Blaschke multi-spectra and a joint spectral segmentation algorithm for modal alignment and spectrum segmentation.
• Employed a voting-based filter bank, informed by gear fault mechanisms, for noise suppression and feature enhancement.

Main Results:

• MBMD effectively integrates multivariate information from gear system vibrations.
• The proposed method demonstrated superior performance in fault feature extraction compared to existing techniques.
• Experimental validation confirmed the effectiveness of MBMD in gear fault diagnosis.

Conclusions:

• MBMD offers a novel and effective approach for mechanical fault diagnosis in gear systems.
• The method's ability to integrate multivariate information and account for mechanical characteristics leads to more accurate fault detection.
• MBMD provides a new perspective for advancing condition monitoring and diagnostics in rotating machinery.