Research And Implementation Of The Track Signal Dectecion Based On DSPbuilder Technology

As a communication carrier between the train and the ground control system,the railroad track signal carries various monitoring and control information that is necessary for train's running,and also it plays a crucial role of ensuring the train's safety and reliability, so real-time monitoring is needed to check out whether its parameters is normal.The railroad track signal is frequency shift keying(FSK) signal and it results from track circuit's low frequency information modulates high frequency carrier,its frequency represents the digital signal which will be transmitted.The main purpose is to find out its center frequency F₀,upside frequency F_h(F₀+△F),underside frequency F₁(F₀-△F),the low frequency modulated signal,the scope,the voltage and the current and so on.According to these parameters we can judge the railroad track communication system's quality,thus we can maintain the send and receive equipments efficiently and timely.Under the premise of deep study of track signal's characteristics and content,this thesis has referenced the advantages and disadvantages of two detecting method that based on the general DSP chip and pulse,has proposed a new signal detection method that associated with the author's research and practice.This method uses the software tool Matlab to design and simulate signal processing algorithm and realizes the corresponding signal processing circuit in the FPGA chip with the convector of DSPbuilder module. Because each of the generic DSP chips' performance indicators and parameters are fixed,so the developing process of platform based on the general DSP chip is repeated in most situations like making circuit board and programming to simulate the model of whole algorithm.But the development of the tests platform with Matlab software and DSPbuilder components avoid the above redundant process,its convenient lays in that algorithms in the Matlab software can be modified without limit and resources in the FPGA chip are possible to reorganize.The experiment has proved that such method has very good feasibility and usability.And the author thinks that all the digital signal processing instead of track signal will get benefit from this method.To ensure the accuracy and implementation of the above methods,this thesis has discussed in detail the digital signal processing theoretical model which is needed to test track signal frequency include sampling,digital filter,FFT spectrum,re-sampling algorithm and so on,at the same time the author has developed a test platform to prove the related theoretical model.These technologies include calling DSPbuilder components to design and simulate sampling,digital filtering,FFT algorithm module in the Matlab/Simulink, restructuring the simulated sub-module as a whole algorithm,translating the designed document into hardware description language(HDL),convecting hardware description language to circuit connection database in the FPGA development and layouting wiring for circuit connection database in software QuartusⅡ,descripting the signal processing circuit for one of embedded-system devices,using SOPCbuilder to design embedded system and mapping the entire system into the FPGA chip,programming the system to achieve the search of side frequency and calculation of current and voltage and management of man-machine interface.This thesis used these theories and technologies to build a test platform and provided an application of using the DSPbuilder components of Matlab to design and realize digital signal processing.We had detected several simulated track signals through this established testing platform,and the result was satisfiable.Majority of the parameters of this platform can meet our requirements about performance index,and this platform can be used in the actual railroad track signal detection after enhancement and melioration.