Modeling Simulation and Diagnosis Analysis of Inter-turn Short Circuits Fault in Generator Rotor

To better apply the recurrent surge oscillograph (RSO) method to diagnose the inter-turn insulation of generator rotor, we investigated the response characteristics of repetitive pulse, the fault location algorithm as well as the severity analysis of inter-turn short circuit fault. Firstly, the ladder equivalent circuit of generator rotor windings and its equivalent circuit number per unit length are determined by the improved vector fitting algorithm, then the time domain response of RSO signal propagating in generator rotor winding is simulated by circuit simulation software. Based on this simulation model, a multiple signal classification (MUSIC) algorithm is used to estimate the wave front of the RSO injected pulse and the characteristic waveform when the fault of various positions and severity are set, and the fault is located by the time difference and the propagation speed of the pulse. Finally, to study the change trend of the fault severity, the calculation is made by changing the position of fault, the length of inter-turn short circuit and the reference. The results show that the improved vector fitting algorithm has the advantages of high operational efficiency and easy application at the field. And the relative location error by using MUSIC algorithm is less than 1.83%, which can meet the requirements of field testing. The research findings also indicate that the feature waveform distortion area can be used to characterize the fault severity with higher reliability. The results show the effectiveness of the proposed method. © 2019, High Voltage Engineering Editorial Department of CEPRI. All right reserved.