| In the context of the country's promotion of military-military integration,the civilianization of radar and other military products has become a research hotspot,and the intelligent transportation system centered on vehicle-borne radar has developed rapidly.Compared to sensors such as optics and infrared,FMCW radar has the advantages of all-weather,all-day and long-range effects.Compared with other frequency bands,millimeter-wave radar has better distance and Doppler resolution performance.Therefore,with today's rapid development of intelligent transportation systems,millimeter-wave radar has become an integral part of auto-assisted driving systems.Although reversing radars are relatively mature,front-facing anti-collision radars are not yet complete because of complex road scenarios.Therefore,this paper has important theoretical significance and practical value.This paper mainly researches the advanced algorithms of target detection and parameter estimation for FMCW automotive radar.In terms of target detection,two-dimensional high-resolution imaging and detection algorithms are mainly researched.For parameter estimation,we mainly research the measurement and estimation algorithms for target distance,velocity,and angle.Firstly,the design of FMCW radar system is studied,the signal model is analyzed,and the distance-Doppler imaging algorithm and target detection algorithm are studied.Through distance-Doppler two-dimensional coherent accumulation imaging,the signal-to-noise ratio of the image is effectively improved,and the target detection performance is improved.Secondly,in terms of distance measurement and speed measurement,the following researches have been carried out from triangle and sawtooth waveforms: 1)Based on symmetric triangular waves,multi-target distance measurement and velocity measurement using the upper and lower sweep frequency domain pairing methods have been studied;2)or sawtooth waves,based on Distance-Doppler 2D image model,we study distance measurement and speed measurement algorithm.For the velocity fuzzy problem,the velocity deblurring method is studied.The phase interferometric technique is used to solve the velocity fuzzy problem.High-precision velocity estimation can be achieved using only single frame data.For the problem of a large number of false alarms in Doppler sidelobes after constant false alarm detection in two-dimensional FFT schemes,a detection scheme based on DBSCAN clustering was proposed and verified by simulation.Then,in terms of angle measurement,the detailed analysis and performance simulation of classic angle-measurement algorithms such as phase-phase,sum-difference beam and subspace estimation in FMCW radar are performed,and the advantages,disadvantages and application scenarios of each algorithm are compared.Aiming at the problem of the degree of freedom in the subspace algorithm,a hybrid estimation algorithm based on the range gate estimation is proposed.For a single range gate with targets,the high-precision angle estimation is ensured using the phase comparasion approach when single target,and the subspace method is used to distinguish multiple targets are used when multiple targets.In addition,too many data frames is needed for the subspace method.Multi-source interpolation algorithm based on single frame data is studied.The non-parametric amplitude phase estimation algorithm(IAA-APES)combined with iterative algorithm is proposed to achieve high precision.Finally,this paper giving the result of the field testing.We study how to solve array multi-channel error and analyzed the imaging performance to verify the effectiveness of the algorithm in this paper. |
PDF Full Text Request