FAULT DIAGNOSIS METHOD FOR SEGMENT-POWERED SUPPLY SYSTEM BASED ON TRANSFORMER

The segment-powered supply system is an important part of linear induction motor. The effective fault diagnosis method is helpful to the normal operation and troubleshooting of the motor. According to the characteristics of the measured three-phase current waveform, a fault diagnosis method of segment-powered supply system combining an algorithm for extracting the amplitude feature sequence and Transformer model is proposed. In addition, the interpretative method of deep learning model was introduced to assist the model to locate anomalies roughly by it-self. The validity and reliability of these methods were verified by the test data of a segment-powered linear induction motor.