Research Of Signal Processin And Target Detection Technology In Train Anti-collision Radar

At present, railway traffic accidents occurred from time to time in China, passengertrain accidents still exist. Collision avoidance radar system has been widely used inautomobile field, except in the application of train anti-collision. The train collisionavoidance radar as auxiliary safety equipment has very high practical value. This paperdiscusses the basic signal processing of collision avoidance radar, including DDCtechnology and high-speed data transmission technology, and the signal detectiontechnique are analyzed as well as DSP programming realization. Through the systemtests data analysis, the ability of the train collision avoidance radar to measure targetsvelocity, distance and angle has been verified.The basic signal processing includes data acquisition, FIR filter, mixing anddecimation. This paper introduces the implementation of basic signal processing inFPGA. It is designed to take full account of the balance between resources andperformance. Four channel DDC task has been realized in the end. For the specialsystem parameters, the mixing module can be simplified. FIR filter achieved by usingXilinx IP core and the resources are saved by multiplexing. Using PLL+DCM methodprovides a stable and reliable clock for each module. The design processing followsbottom-up, modular thought, then applies the Verilog language to instantiate eachmodule in ISE. Finally, through Modelsim simulation, and the combination of Matlabsimulation, each module function is verified.This paper introduces how to achieve tasks scheduling and resources allocation byusing of SYS/BIOS operating system on DSP. The DSP finished the transmission ofecho data through uPP interface, and then through the echo signal processing to detectobstacles, finally using NDK realized data transfer from DSP to the PC by Ethernet.Through the basic signal processing test and system debugging, this paper obtaineda preliminary understanding of system performance. And this paper did data acquisitionin the train actual operation environment, for the detection of various track obstacles.According to the analysis of experimental data, this paper proposed some correspondingdetection method. The final results show that the detection is basically right and thesystem is workable.