Experiment research and analysis of stator winding short circuit faults in DFIG

Abstract

Doubly-Fed Induction Generator (DFIG) plays a significant role in wind power industry. It usually works in harsh environments. According to reliable surveys, 38% of DFIG failures are related to the winding faults and huge cost is required for maintenance. Therefore the condition monitoring system is of great importance for wind farms. In this paper, an experimental platform has been setup to detect stator winding faults, such as inter-turn, phase to phase and phase to ground short circuit fault. Some fault feature parameters based on stator line current which are useful for fault diagnosis are extracted and analyzed. The RMS, phase difference, negative-sequence, and the Park's vector trajectory's eccentricity are extracted as feature parameters. Stator line currents are balanced under normal condition, so the negative sequence of stator currents is almost zero, and their phase differences are about 120°. The park's vector trajectory is a circle under healthy condition, and its eccentricity is near to zero. The ratio of maximum and minimum RMS(root mean square) shows difference when different fault types occur, the value of the ratio is larger when phase-phase short circuit occurs compared with inter-turn and phase-ground short circuit faults; the eccentricity of Park's Vector trajectory would increase with the fault degree severity. Besides, negative-sequence component gives useful information, it measures the imbalance of the three-phase currents, the negative-sequence component in the stator current is increasing as the fault level grows. These feature parameters' variation trend shows that phase to phase short circuit is the most severe fault, and phase to ground short circuit is more serious fault than inter-turn short circuit. Different faults can be distinguished through these fault features. This will contribute to fault diagnosis of DFIG.