Application of wavelet denoise in the harmonic analysis of electrical traction power supply system

Electric locomotive is the main unbalance load and harmonic source load of power system, whose typical harmonic current mainly are 3, 5, 7 times high frequency harmonic currents. These high frequency harmonic currents will do great harm to the safe and stable operation of electric locomotive. Taking type SS8 electric locomotive as an example, the working process of its rectifying circuits were simulated, and the harmonic current characteristics of the electric traction power supply system were analyzed. Wavelet decomposition, threshold disposal and wavelet reconstruction were carried through on the harmonic current signal of electric traction power supply system for the purpose of denoising its harmonic currents. In order to search the best wavelet threshold for the better denoise of the harmonic signal in electric traction power supply system, adaptive threshold selection based on Stein unbiased likelihood estimation theory, heuristic threshold selection, sqrt (2log(length(X))) threshold selection and minimax threshold selection were adopted in simulation. Various wavelet denoise thresholds were used to reduce and eliminate high frequency harmonic current content of signal according to different characteristics of harmonic signal. Simulation results show that the best threshold of the harmonic current signal in electric traction power supply system is sqrt (2log(length(X))) and it can reduce high frequency harmonic currents of signal obviously after its de-noising.