Fault Diagnosis Method for Vacuum Contactor Based on Time-Frequency Graph Optimization Technique and ShuffleNetV2

Area of Science:

• Electrical Engineering
• Signal Processing
• Machine Learning

Background:

• Vacuum contactors are critical in power systems, but their fault diagnosis faces challenges due to inefficient feature extraction and redundant data.
• Existing methods struggle with low diagnostic performance, necessitating improved techniques for reliable operation.

Purpose of the Study:

• To develop a high-performance fault diagnosis method for vacuum contactors.
• To address limitations in feature extraction and data redundancy in current diagnostic approaches.

Main Methods:

• Vibration signals from vacuum contactors were transformed into generalized Stockwell transform (GST) time-frequency graphs.
• Multi-resolution GST graphs were generated, and the OTSU algorithm was used for energy concentration area cropping, optimizing graph size by 68.86%.
• The ShuffleNetV2 network was employed for enhanced feature learning and fault classification.

Main Results:

• The proposed method achieved a fault recognition accuracy of 99.74% for the CKJ5-400/1140 vacuum contactor.
• Single iteration time for model training was reduced by 19.42%, indicating improved efficiency.
• The optimized time-frequency graph and ShuffleNetV2 integration effectively improved diagnostic performance.

Conclusions:

• The GST-based time-frequency graph optimization combined with ShuffleNetV2 offers a highly accurate and efficient fault diagnosis solution for vacuum contactors.
• This method effectively overcomes the limitations of inadequate feature extraction and data redundancy, paving the way for more reliable electrical equipment monitoring.