Incipient Interturn Short-Circuit Fault Diagnosis of Permanent Magnet Synchronous Motors Based on the Data-Driven Digital Twin Model

As the most common fault of permanent magnet synchronous motor (PMSM), interturn short-circuit fault (ISCF) has great harm and develops rapidly. Once it is not diagnosed in time, it will bring secondary damage to the motor system. In the process of early fault diagnosis, the harmonic of the motor often increases the difficulty of fault feature extraction. In order to improve the reliability of the system, a method for early interturn short-circuit diagnosis of PMSMs based on data-driven digital twin models is proposed in this article. First, the three-phase current residuals of the PMSM under ISCF are analyzed theoretically. Second, the digital twin model of the target motor in healthy states is established through nonlinear auto-regressive model with exogenous inputs (NARX) network. Finally, the early short-circuit diagnosis is completed through the analysis of the three-phase current residual. The method proposed in this article does not need fault data and can complete the diagnosis of minor faults under the condition of large harmonic and certain inherent asymmetry. It has high sensitivity and obvious fault characteristics. Theoretical derivation and experiments verify the effectiveness of the proposed method.