A Novel Detection Scheme for Motor Bearing Structure Defects in a High-Speed Train Using Stator Current

Area of Science:

• Mechanical Engineering
• Electrical Engineering
• Condition Monitoring

Background:

• Railway traction motor bearings (RTMB) are vital for high-speed trains (HST) but prone to failure.
• Existing monitoring systems suffer from high false-positive rates.

Purpose of the Study:

• To introduce a novel sensorless monitoring scheme for RTMB using stator current.
• To improve the accuracy and efficiency of fault detection in RTMB.

Main Methods:

• A hybrid signal preprocessing algorithm combining adaptive notch filtering (ANF) and envelope spectrum analysis (ESA).
• Utilizing a deep belief network (DBN) for health status classification.
• Experimental validation on a 1:1 scale high-speed railway traction motor.

Main Results:

• Mechanical defects in RTMB can be reliably detected through stator current changes.
• The proposed sensorless method achieves at least 17.3% higher fault detection accuracy than vibration-based methods.
• The developed scheme demonstrates superior accuracy (at least 8.9% higher) in diagnosing fault presence and type compared to other methods.

Conclusions:

• The novel sensorless monitoring scheme effectively detects RTMB faults using stator current.
• This approach offers a more accurate and efficient alternative to current RTMB monitoring practices.
• The study validates the feasibility and effectiveness of stator current analysis for RTMB condition monitoring.