SF₆ High-Voltage Circuit Breaker Contact Status Detection at Different Currents

Area of Science:

• Electrical Engineering
• Materials Science
• Signal Processing

Background:

• Online non-destructive testing (NDT) methods for high-voltage circuit breaker (HVCB) contact states, especially those using SF6 gas, are currently limited.
• Accurate monitoring of HVCB contact states is crucial for reliable power system operation and fault prevention.

Purpose of the Study:

• To develop a novel method for detecting the contact state of HVCBs using vibrational signals.
• To establish a reliable and accurate HVCB contact state recognition model for online monitoring.

Main Methods:

• Designed a mechanical vibration detection system and high-current generator for a 40.5-kV SF6 HVCB prototype.
• Collected vibration signal data under various current conditions and corresponding contact states.
• Employed frequency-based feature extraction, optimized deep neural networks (DNNs) with Adam, and Bayesian optimization (BO) for model tuning.

Main Results:

• The proposed recognition model accurately identified five distinct HVCB contact states.
• The model demonstrated high recognition accuracy (above 96%) for currents ranging from 1000 A to 3500 A.
• Validated the effectiveness of the vibrational signal analysis and DNN approach for HVCB monitoring.

Conclusions:

• The developed vibrational signal-based method offers a promising non-destructive approach for HVCB contact state monitoring.
• The study provides a foundation for future research in HVCB fault diagnosis and condition monitoring.
• The optimized DNN model enables accurate and reliable assessment of HVCB performance under various operating currents.