Abstract: In view of the difficulty in identifying the generator stator winding interturn short circuit fault accurately, the dynamic feature identification method based on PSCAD/EMTDC simulation was studied. By constructing a 3D fault model of generator in PSCAD/EMTDC simulation environment, the electromagnetic transient process of interturn short-circuit fault is simulated. On this basis, the dynamic spectrum characteristics of the fault current are extracted, which can sensitively reflect the change of small electrical volume caused by the interturn short circuit. Further, a millisecond fault detection method is designed to realize real-time identification of interturn short circuit faults by using dynamic spectrum characteristics. Compared with the existing methods, the proposed method can significantly improve the accuracy of fault detection, and can effectively identify faults at the initial stage, which provides a strong technical support for the timely maintenance and safe operation of generators.