Identification of inrush currents in power transformers based on higher-order statistics

This paper presents a novel method based on higher-order statistics for discrimination between magnetizing inrush and internal fault in power transformers. The distribution characteristics of the sample points of differential currents, under various operation conditions, are firstly investigated and quantified using higher-order statistics. Based on these characteristics, three kurtosis-based indices are defined for distinguishing inrush current from internal fault current. The final discrimination criterion combines these three indices to improve the performance of inrush current identification. Extensive simulation studies and experimental tests verify the effectiveness of the proposed method and its advantages over the conventional second harmonic restraint method.