Harmonic detection in traction power system based on improved Hilbert-Huang transform

The improved Hilbert-Huang transform is applied in harmonic detection in traction power system, in this paper. Because the energy of fundamental is much larger than the other harmonics´, mode-mixing appears during the sifting by Empirical Mode Decomposition (EMD), and consequently, each single harmonic component cannot be effectively extracted from the complex original signal. In order to solve the mode-mixing problem, the improved method, EMD based on Fourier transform, is used in this paper. The improved HHT method is applied to analyse harmonic voltage and harmonic current in traction power system, and then complex harmonic signal can be decomposed into single harmonic component, and time-amplitude and time-frequency of harmonics can be obtained accurately. The results show that the improved method has better performance in harmonic analysis than wavelet packet analysis in traction power system.