A Study on Machine Learning-Based Feature Classification for the Early Diagnosis of Blade Rubbing

Area of Science:

• Mechanical Engineering
• Artificial Intelligence
• Condition Monitoring

Background:

• Blade rubbing is a critical fault in rotary machinery, leading to significant damage and operational downtime.
• Early detection of blade rubbing is essential for preventing catastrophic failures and ensuring machinery reliability.

Purpose of the Study:

• To develop and validate a machine learning-based approach for the early diagnosis of blade rubbing in rotary machinery.
• To differentiate between rotor unbalance and blade rubbing faults using signal processing and feature analysis.

Main Methods:

• Simulated blade rubbing and rotor unbalance faults using experimental models.
• Applied machine learning diagnostic methods to analyze vibration signals.
• Utilized signal processing and feature analysis for fault diagnosis.

Main Results:

• Successfully distinguished between unbalance and blade rubbing based on distinct signal trends.
• Demonstrated the capability to identify blade rubbing before it escalates.
• Showcased the effectiveness of machine learning in diagnosing varying severities of blade rubbing.

Conclusions:

• The developed machine learning approach enables early diagnosis of blade rubbing in rotary machinery.
• This method provides a reliable way to differentiate unbalance from blade rubbing faults.
• The findings support the application of machine learning for proactive condition monitoring and fault prevention in rotating equipment.