The Research Of Signal Processing Algorithm Of Doppler Speed Measurement Sensor

In recent years, China’s railway transport, especially the high-speed rail growth hasbeen accelerated. The railway department for rail transportation operation carried outseveral large scales of accelerations for the speed of the train. Considering this background,real-time speed measurement of the high-speed trains put forward higher requirements totraditional velocimetry. Based on the principle of the Doppler, the radar system formeasuring the key parameters of high-speed rail operating speed, mileage and directionhas a higher accuracy and timeliness. The anti-jamming capability of the radar systemunder a variety of terrains, climate and environment is also stronger. The Doppler velocitysignals collected by velocimetry system rely mainly on its digital signal processor powerto estimate the spectrum, and then the system can calculate the accurate speedinformation.By the beginning of this thesis, it elaborated the radar speed train technology basedon the Doppler effect, focusing on the study of train speed sensor Doppler signalprocessing algorithms. Then this thesis took the railway speed sensing requirements fromthe current railway department into account, and analyzed the feasibility of the systemand the corresponding DSP program algorithm.This thesis compared the two different spectral estimation algorithms which are Burgalgorithm of AR model parameters and FFT algorithm and determined the choice of FFTFast Fourier Transform spectral estimation algorithm as the core algorithm for signalprocessing of this radar velocimetry system. Using MATLAB simulation to FFTalgorithm, this thesis analyzed the influence of the choice of FFT parameters on the resultsof spectral estimation, and determined the selection principle of parameters. Meanwhile,an effectively improved method of FFT algorithm was put forward to simplify thecalculation steps. The Matlab software simulation results further approved the efficiencyof the improved FFT algorithm. Then this thesis expanded the radar sensor system for thebasic performance test, and the test results were analyzed. Test results showed that theselected spectral estimation algorithm can substantially meet the system designrequirements.Study of this system for train speed plays a role in the promotion of FFT algorithm insignal processing, and also has theoretical and practical significance on our currentresearch in the field of rail transport speed.