Research on Intelligent Identification of Magnetizing Inrush Current based on Empirical Modal Decomposition

Abstract The accurate identification of the converter transformer magnetizing inrush current can ensure safety operation UHV DC transmission system. In differential protection transformer, it is extremely hard to distinguish between excitation generated when closed and Fault due internal fault, which leads frequent misoperation extent that access fails. This paper proposes a model based on empirical modal decomposition (EMD) probabilistic neural network (PNN) algorithm with particle swarm (PSO) optimization. EMD decomposes waveform, eigenmode functions (IMFs) three highest correlation bars are selected. energy, cliffs, standard deviation IMFs extracted as original waveform eigenvectors.