Research On The Application Of Wavelet Transform In Earthquake Disaster Prevention System Of Traction Power Symtem For High-Speed Pailway

China has a vast territory and complicated geological conditions, and our country is one of the countries where earthquakes often take place in the world. Nowadays, because of the rapid development of high-speed railway, security has become the the important task of the railway transportation. The damage caused by the earthquakes is inestimable, especially for high-speed train trains.We can not predict earthquakes with the present technology, so as to reduce the hazards of earthquakes on the high-speed rail, so it is necessary to establish the system of earthquake early warning in order to reduce the hazards of earthquakes on the high-speed rail.According to the earthquake early warning systems for high-speed railway in Japan and Franceand, this paper will study on the construction of earthquake early warning system high-speed Rail in china, combined with the actual situation of the existing Chinese. Earthquake early warning system is a kind of alarm system based on P seismic monitoring.The system will estimate the seismic parameters by P-wave signal and calculate the seismic acceleration along the high-speed rail automatically, and then determine the need for control of trains. Main points in development of the earthquake early warning system for high-speed railways are discussed, including threshold for early warning, layout of monitoring stations, train operation and structure of the early warning system.Earthquake early warning system for high-speed railway can monitor the seismic information real-time by installing seismic instruments along the railway lines and in potential seismic source zone.Because the earthquake information contains a lot of noise,it will greatly enhance the possibility of false alarms though processing the signal containing noise directly.Therefore, choosing a suitable denoising method is very important.The effective signals and the noise of seismic information have some overlap in the frequency domain, Therefore, it can’t remove the noise only by the frequency domain filtering. As a time-frequency analysis method, wavelet transform can arbitrarily change the spatial resolution of signal and decompose the signal multi-scale, so it can be better to target signal analysis. This paper mainly studies the application of wavelet transform technology in the earthquake early warning systems for high-speed railway, and the construction of denoising the target signal using the new threshold function.The Ricker wave, EL-CENTRO wave and Japan’s Kobe earthquake wave will be denoised by the method of Wavelet threshold denoising in the MATLAB software simulation. The experimental results show that:threshold denoising method of referenced threshold function for suppressing noise has obvious advantages;P-wave and S-wave will be separated obviously through the signal denoising; The system can test the P waves and greatly reduce the probability of false alarm using the denoised signals.