Diagnosis of Stator Inter-Turn Short Circuit Faults in Synchronous Machines Based on SFRA and MTST

As a key component of the power system, the good or bad conditions of synchronous machines will directly affect the stable supply of electric energy. The inter-turn short fault of the stator is one of the main dangers to the synchronous machine and is difficult to diagnose. Frequency response analysis has recently been introduced and used for detecting this type of fault; however, the fault degrees and locations cannot be directly recognized by traditional frequency response analysis. Therefore, this study improves the frequency response analysis by combining it with a deep learning model of a multivariate time series transformer. First, the principle of this study is introduced. Second, the frequency response data of short circuit faults are obtained using an artificially simulated experimental platform of a synchronous machine. The deep learning model is then well-trained. Finally, the performance of the proposed method is tested and verified. It concludes that the proposed method has the potential for classifying and diagnosing the inter-turn short circuit of stators in synchronous machines.