Study On Electrified Railway Traction Loads Harmonic Analysis Method Based On Wavelet Transform

Recent years, electrified railway of China develops quickly, which not only makes the traffic convenient but also produces enormous economic benefits. As the main traction load, electric locomotives are driven by electricity, which shows its predominance of high efficiency, energy conservation and environment protection compared with diesel locomotives. But the main circuit of electric locomotives composes of many power electronic devices and the locomotives'running condition are complicate, which make electrified railway injects a lot of harmonics into public power network with gaining power frequency energy from power grid. Because the harm of harmonics is serious and it has caused many major incidents in European and American, we must take preventive measures and make a good job in harmonic research of electrified railway traction loads.Commonly used methods of the harmonic analysis include FFT, the nerve network, wavelet transformation, harmonics analysis based on the extend Prony algorithm and so on. Wavelet analysis is a hot topic of harmonic analysis and it is also the new way and new theory currently. This paper applies wavelet transformation to the harmonic analysis of the electrified railway traction load, and verifies that the necessity of wavelet function selection by comparing the analysis results of the same harmonic by different function. The best wavelet function db10 which is used in harmonic analysis is defined by MDL principle, the reasonable decomposition level is determined. Finally the harmonic analysis of HXD30017 locomotive currents base on MATLAB is realized by running wavelet algorithm codes which are programmed by functions in wave toolbox, and the valid values of harmonic currents are calculated by using the theory of wavelet packet analysis. At the same time, the Fourier analysis on harmonic currents is realized and signal's spectrum is obtained. The comparisons between the simulated results of two transform methods show that wavelet transform is better than FFT.The simulation results express that wavelet transform can decompose signals to fundamental wave and different frequency harmonic waves exactly, so completes harmonic detection. We confirmed that wavelet transform can bring a more exact analysis of the local characteristics of signals, the transient signal can be analyzed correctly, so as the real-time and precision of harmonic measurement can be improved too. In addition, we can get the relevant information of harmonics rapidly and conveniently by wavelet transformation, such as valid values. Therefore, wavelet transformation can be applied in the harmonic analysis of the electrified railway traction loads.