Envelope analysis with a genetic algorithm-based adaptive filter bank for bearing fault detection.

Area of Science:

• Mechanical Engineering
• Signal Processing
• Condition Monitoring

Background:

• Early detection of bearing defects is crucial for preventing equipment failure.
• Envelope analysis is a common technique for bearing fault detection.
• Optimal filter band selection remains a challenge in envelope analysis.

Purpose of the Study:

• To develop a robust fault detection methodology for bearings.
• To optimize the passband selection for envelope analysis using a genetic algorithm.
• To improve the reliability of bearing defect identification.

Main Methods:

• Utilized envelope analysis combined with a genetic algorithm (GA)-based adaptive filter bank.
• Explored the impact of various passbands specified by the GA.
• Evaluated passband performance using a residual frequency components-to-defect frequency components ratio.

Main Results:

• The genetic algorithm successfully identified an optimal passband for bearing fault detection.
• The proposed method demonstrated improved reliability in identifying bearing defects.
• The ratio metric effectively quantified the degree of bearing defectiveness.

Conclusions:

• The GA-based adaptive filter bank provides an effective solution for optimal passband selection in bearing fault detection.
• This methodology enhances the early and reliable identification of bearing defects.
• The study contributes to advancing condition monitoring techniques for rotating machinery.