Fault Diagnosis of High-Voltage Circuit Breaker Based on Digital Twin

Abstract

High-voltage circuit breaker is one of the key components in the power grid, and mechanical faults often occur during long-time operation. A method based on digital twin for mechanical fault diagnosis of high-voltage circuit breaker is proposed in this paper. First, a finite element twin model is established to simulate the opening and closing operation of high-voltage circuit breaker to simulate its typical coil currents, then measured coil currents are obtained from the actual circuit breaker. Both the virtual data and measured data are utilized to train the KNN model to perform the fault diagnosis of five operation states. Experimental results show that the proposed digital twin method can simulate the operation of high-voltage circuit breaker, provide more failure types of data and can efficaciously diagnose mechanical faults with a high classification accuracy.