Real-Time Evaluation of Ground Insulation Degradation and Fault Warning Method Under Multiple Operating Conditions...

Area of Science:

• Electrical Engineering
• Railway Transportation Systems
• Fault Diagnosis and Monitoring

Background:

• Existing methods for diagnosing main circuit grounding faults in locomotive and high-speed train traction drive systems are limited to single operating conditions.
• Early warning and comprehensive fault diagnosis across all operational states are crucial for ensuring train safety and preventing cascading failures.
• Ground insulation degradation poses a significant risk, necessitating advanced monitoring and warning systems.

Purpose of the Study:

• To propose a mechanism and data-driven real-time evaluation method for early warning of ground insulation degradation in traction drive systems.
• To enable fault diagnosis and warning under all operating conditions, overcoming the limitations of existing single-condition diagnostic approaches.
• To enhance train safety and support optimal degraded operation strategies through accurate fault information.

Main Methods:

• Analysis of circuit characteristics and mathematical modeling of grounding faults and sensor signals under various operating conditions.
• Extraction of a feature index set for accurate classification of different grounding fault types, informed by engineering experience.
• Development of a fault diagnosis rule library using a decision tree classification model trained on on-site fault data and real-time sensor signal collection.

Main Results:

• Accurate detection and localization of grounding faults across diverse operating conditions were achieved.
• Real-time warning of the severity of grounding faults was successfully implemented.
• Controller hardware-in-the-loop testing validated the method's effectiveness and engineering applicability.

Conclusions:

• The proposed method provides a robust solution for real-time, full operating condition grounding fault detection and warning in train traction drive systems.
• The system ensures enhanced train safety by enabling early identification of insulation degradation.
• The approach offers valuable data for future optimal degraded operation planning.