| With the rapid development of autonomous vehicle technology,the demand for environmental awareness has become even higher.Millimeter wave radar plays an important role in autonomous driving systems because of its advantages of low cost,good motion target detection capability and adaptability to various weather conditions among commonly used automotive radars.In this thesis,the signal processing technology of automotive radar is studied based on 77 GHz linear frequency modulated continuous wave(LFMCW)automotive radar,a set of automotive radar signal processing process design scheme is proposed,and simulation design and actual scenario testing are carried out,the specific work of this thesis is as follows:1.The basic principle of LFMCW radar measuring distance,speed and angle is introduced,the basic structure of LFMCW radar system is explained,and the components of the system and the functions of each part are described.2.To address the problem that the spectral leakage effect of the differential beat signal causes errors in target estimation,more accurate distance and velocity information is obtained by adding time-domain window functions before the distancedimensional FFT and velocity-dimensional FFT processing,respectively.For the case of excessive transient noise due to complex background during detection,the noise is averaged and the detection sensitivity is improved by performing incoherent accumulation.In order to make the automotive radar detect the target more accurately in complex scenes,two constant false alarm rate(CFAR)detection techniques commonly used in automotive radar are studied,and good results are achieved in the simulation and real-world scenario detection applications.3.For the problem of low performance of radar angle FFT angle measurement method,the design combines the principle of time division multiplexing multiple input multiple output(TDM-MIMO)radar extended receive array antenna aperture and proposes the method of putting the multiple signal classification(MUSIC)algorithm after CFAR detection for the case that the automotive radar may detect more targets than the number of arrays at the same time.This method of angular super-resolution only for the detected points avoids the contradiction between the number of targets and the number of arrays and improves the radar angular resolution significantly,from 14.5° to 5.0° in the simulation results.4.The design of the signal processing process is completed and the rationality of the whole process is verified in the form of MATLAB simulation,and the road is measured based on the AWR1642 millimeter wave radar module,and the processing results of the radar differential shot signal are compared with the experimental scenario to verify the correctness of the signal processing process proposed in this thesis.The thesis contains 61 figures,6 tables,58 references. |
PDF Full Text Request