Identification of the Series Arc Fault Based on EMD Algorithm

An improved control theory arc model is proposed to control micro arc distance between two electric poles. An arc voltage wave is measured on an arc generation device to simulate serial arc fault in 220Vac-50Hz low-voltage distribution network. Arc identification based on empirical mode decomposition (EMD) algorithm is achieved to analyze the intrinsic mode functions (IMFs) under normal and fault condition. EMD decomposition under normal arcs and fault arcs is achieved to obtain a series of intrinsic mode components, which is beneficial to finish the identification of fault arcs. The maximum value of the IMF9 mode obtained by EMD decomposition can be used as one of the characteristic quantities for identifying fault arcs.