Research Of Orbit Signal Detecting Using Wavelets Transformation Theory

There are more and more electrified railways in rail transport systems with the rapid growth of China's railway industry, which has brought great convenience for people's traveling. And with the speed of train running higher and higher, new ways to detect orbital signal are required.In order to meet the needs of rapid development of the railway, it's necessary to study a reliable, accurate and a good anti-jamming detecting method to analyze railway signal in real-time.At present, there are many formats of track circuit in the electrified section in China, such as 25 Hz AC counting signal,75 Hz AC counting signal, frequency-shift keying track circuits.As the track circuit of 25 Hz AC counting signal and ZPW-2000A non-insulated frequency-shift keying track circuit is used widely in electrified section, this paper take them as example.As there are various sources of interference in the field environment, this paper firstly analyze these two signals both in time domain and frequency domain, secondly use the theory of wavelets transformation theory to de-noise the noise-signal,finally analyze the signal of 25 Hz AC counting signal and frequency-shift keying track circuits.For 25 Hz AC counting signal, this paper firstly deal with the noise using the threshold de-noising method, including the ways of a hard threshold value, soft threshold value and the half-soft threshold value method, and at last the Matlab simulation results were given. At the demodulation part, this paper use the method of wavelets transform theory to detect the ideal 25 Hz AC counting signal and noisy 25 Hz AC counting signal.For ZPW-2000A non-insulated frequency-shift keying signal, this paper describes how to collect the data from ZPW-2000A non-insulated frequency-shift keying track circuit. At the part of de-noising for frequency-shift keying signal, this paper taking the methods of wavelets denoising and wavelet packet de-noising. From the simulation result, it get the conclusion that the noise signal was filtered out, which verified that the wavelets and wavelet packet's superiority in de-noising. For the part of demodulating and detecting the collected frequency-shift keying signal data, this paper using some regular kinds of wavelets algorithms to decompose the signal into five level, such as db4,db5,coif5 wavelets.From the result of simulation experiments, it can see that using wavelets theory detecting 25 Hz AC counting signal has the characteristics of strong ability of resisting disturbing and small error. In addition, it can effectively overcomes the shortcoming of the traditional method, such as 0,1 threshold is hard to select; For ZPW-2000A non-insulated frequency-shift keying signal demodulation, wavelets algorithm can effectively improve the detecting accuracy and frequency resolution, which have an important practical significance and practical value. The way of using wavelets transformation theory detecting signal can be an effective complement to the current detecting methods.