A New Method to Identify Inrush Current Based on HHT

Wrong operation of transformer protection is always caused by magnetizing inrush currents. For safe running of a transformer, it is necessary to discriminate between magnetizing inrush currents and internal faults. Because the inrush current is a peaked wave with nonstationary component, the internal fault current can be distinguished from inrush current by the energy characteristics with time and frequency. A new method based on Hilbert-Huang transformation (HHT) was developed. Firstly, empirical mode decomposition (EMD) separated the electric current series to components with different time scale, say, intrinsic mode function (IMF). Secondly, Hilbert transform was applied to every IMF. Then, three dimensional amplitude (energy)-frequency-time spectrum designated as Hilbert spectrum was got, where both amplitude and frequency were functions of time. The Hilbert spectrum provided high-resolution time-frequency characteristics of transformer´s different work states. The results show energy characteristics of current series. It is demonstrated by actual analysis that the new discriminating method is one effective way for power transformer protection