Band Relevance Factor (BRF): A novel automatic frequency band selection method based on vibration analysis for...

Area of Science:

• Mechanical Engineering
• Signal Processing
• Condition Monitoring

Background:

• Effective monitoring of rotating machinery is critical for industrial production.
• Extracting relevant information from vibration signals is key for condition monitoring.
• Existing methods for frequency band selection often miss crucial information and require manual parameter tuning.

Purpose of the Study:

• To propose an automatic approach for selecting all relevant frequency bands in vibration signals.
• To overcome limitations of existing methods that focus only on impulsive excitations or require manual adjustments.
• To identify frequency bands indicative of changes in machine dynamics or developing faults.

Main Methods:

• Development of a novel method named Band Relevance Factor (BRF).
• BRF utilizes spectral entropy for automatic identification of relevant frequency bands.
• Results are presented via a relevance ranking and heatmap visualization.

Main Results:

• BRF successfully performs automatic selection of relevant frequency bands.
• The approach identifies bands related to machine behavior changes and faults.
• Validation on synthetic and real-world datasets confirms BRF's effectiveness.

Conclusions:

• BRF offers an effective, automatic solution for vibration signal analysis in rotating machinery.
• The method enhances the ability to detect faults and monitor machine health.
• BRF provides valuable insights into machine condition by identifying all relevant spectral regions.