Research on Establishing Data-driven Diagnosis System of Transformer Winding’s Short-circuit Withstand Ability

Power transformers will suffer short-circuit shocks during operation. Multiple impacts will cause irreversible damage to the windings. This will gradually reduce the short-circuit withstand ability of the windings, and will eventually cause the transformer to lose stability and shutdown. In order to discover this problem in time, this paper proposes a method of establishing a data-driven diagnosis system for winding short-circuit withstand ability. First, the voltage, current and vibration signals of the transformer are transmitted to the cloud server platform through the Internet of Things technology; Secondly, obtain the current state characteristic information of the transformer through cloud computing; Repeat the above process to collect a large amount of data about the normal operation and fault operation status of the transformer to construct a Random Forest diagnosis model. Finally, input the characteristic data at the current moment, and the diagnostic model automatically decides whether the transformer is in a winding fault state. Experiments show that the diagnostic system has excellent evaluation accuracy, and the system is practical.