Design Of FMCW Radar Control And Information Real-time Processing Unit

Frequency Modulated Continuous Wave(FMCW)radar has been widely used in industrial level measurement,traffic detection and other fields due to its advantages of high bandwidth,narrow beam and good directivity.In the signal processing of FMCW radar,the traditional high-resolution frequency measurement method based on FFT has greater requirements on the performance of the signal processing system,which increases the hardware cost of the radar system.Therefore,studying high-resolution,fast signal processing algorithms and implementation methods suitable for low-cost microcontrollers not only has theoretical significance for frequency analysis technology,but also can significantly reduce the cost of FMCW radar systems.The thesis first analyzes the measurement principle of FMCW radar and the key factors that affect the accuracy of FMCW radar target measurement,and then uses the 120 GHz radar transceiver front-end chip,and designs for the parasitic amplitude modulation of the radar front-end voltage-controlled oscillator and the characteristics of the FMCW radar difference frequency signal.The frequency source of the modulated signal and the filter amplification circuit are finally realized,and the real-time signal processing unit based on STM32F303 is finally realized.Secondly,to solve the problem of low ranging accuracy when using the Fast Fourier Transform(FFT)method to analyze radar difference frequency signals,a local refined high-resolution frequency estimation algorithm based on Fourier transform is used to reduce Based on the original algorithm,based on the original algorithm,the original algorithm was improved to solve the problem that the original algorithm has a large amount of calculation and the error increases when the real frequency is near the frequency spectrum line.Thesimulation results show that the improved algorithm achieves high-resolution frequency estimation of the difference frequency signal,solves the problem that the original algorithm increases the error when the frequency deviation is small,and has good anti-noise performance on the basis of reducing the calculation amount.When the relative frequency deviation is any value,the improved algorithm's root mean square error of frequency estimation is close to the Cramer-Rao Lower Bound(CRLB)of real sinusoidal signal frequency estimation,which has good consistency and is refined.When the multiple is higher,the algorithm performance is higher than the Quinn algorithm and Candan algorithm based on frequency interpolation.In terms of algorithm implementation,the paper designs and implements a set of feasible signal processing schemes based on the low-cost STM32 platform,and improves real-time performance on the basis of high-precision ranging.According to the scheme proposed in the paper,based on the FMCW radar experimental prototype,the ranging experiment was carried out in the laboratory environment,and the improved algorithm was verified.Finally,some experimental measurement results are given in this paper.Experimental data shows that in the range of 1-10 m,the range error is less than5 mm on average,which meets the expected requirements.