Research On Demodulation Method Of Balise Uplink Signal Based On Duffing Oscillator

With the increase of train running speed and railway transport density, the requirementfor security and reliability of train control system is set higher. Only when train operationcontrol system transmits real-time reliable information between the train and ground canensure safe operation of trains. Balise uplink signal acts as important information that groundbalise sends to vehicle equipment, its real-time demodulation and decoding accuracy is aprerequisite to achieve traffic safety. Since the railway environment is bad, balise uplinksignal during transmission is not only affected by the strong interference of the downlinksignal, but also by other environmental noise, so the balise uplink signal that the vehicleequipment receives is relatively weak. Currently, balise uplink signal detection methods aretrying to detect the useful signal on the basis of filtering noise. These methods remove noise,while the useful signal is weakened, and the detection accuracy is largely dependent on theperformance of the filter, the detection accuracy is reduced with the noise increasing.On the basis of the excellent features of Duffing oscillator system’s sensitivity to theweak signal and noise immunity, the Duffing oscillator detection system is applied to thebalise uplink signal demodulation. With Matlab software as a simulation platform to establisha double demodulation Duffing oscillator model, it is achieved demodulation simulation ofbalise uplink signal.Firstly, the thesis elaborates the principle of balise system, analyzes the characteristics ofthe balise uplink signal, and the current balise uplink signal demodulation method. Based onthe characteristics of the balise uplink signal, and the principle of direct digital frequencysynthesis, the modulation model of balise uplink signal is established.Secondly, the thesis introduces the dynamics characteristics of the Duffing oscillator,improves the general model of Duffing equation, so that it can detect any frequency signal,and then analyzes the principle that the improved Duffing oscillator system detects weaksinusoidal signals. Based on the description for the generate phase diagram state identificationmethods of chaotic system, the thesis analyzes the advantages and disadvantages of variousidentification methods. The Lyapunov exponent discrimination is mainly studied, and theLyapunov exponent calculation methods of the known kinetic equation and of the konwn timesequences are analyzed.Finally, in combination with Duffing oscillator phase change conditions and baliseuplink signal characteristics, methods and procedures are given to demodulate Duffingoscillator in balise uplink signal. Using Matlab/Simulink simulation software to build doubledemodulation Duffing oscillator model, respectively under no noise background,27.095MHzdownlink signal interference, Gaussian white noise and colored noise background Duffing oscillator demodulation in balise uplink signal is experimental simulated and analyzed, andcompares with the traditional demodulation method error rate. Experimental results show thatDuffing oscillator demodulation method in the balise uplink signal is feasible, and the methodhas high detection accuracy and excellent noise immunity.