Research On Multi-target Detection And Tracking Method Of Millimeter Wave Automobile Collision Avoidance Radar

Millimeter wave automotive radar plays an important role in manless driving and application of advanced driver assistance systems in active anti-collision.This subject is aimed at the research of multi-target detecting and tracking in realistic road conditions.The main tasks include:1.In order to tackle the problem of multi-target detecting of millimeter wave automotive radar,one type of multi-slope frequency modulation waveform is employed.Making use of two sets of symmetric triangular wave of different periods,it forms four segments of linear frequency modulation waveform with different frequency modulation slopes for detecting multiple targets.Combined with the corresponding target matching methods,this waveform can extract information of distance and speed of multiple targets,solving the problem of distance-velocity coupling and ghost target of traditional waveform and signal processing methods.Off-site tests indicate that multi-slope frequency modulation waveform and its signal processing methods can detect multiple targets efficiently.2.In order to tackle the problem of target detection in the environment of road clutter,ordered statistics constant false alarm rate(OS-CFAR)algorithm is used that performs well in multi-target situations.By doing so,the target detection of automotive radar has constant false alarm rate and the capability of anti-multi-target interference.Simulation results and experiments demonstrate the effectiveness of OS-CFAR algorithm.3.In order to tackle the problem of spectral broadening of beat signal which is caused by vehicle multi-scatter points,cluster analysis is adopted to classify multi-scatter points that belong to the same target in frequency domain after detection.In order to tackle the problem of frequency measurement error of beat signal’s frequency caused by FFT"Picket Fence Effect",The energy centrobaric algorithm is used to correcting the spectrum so as to extract high-precision parameters of distance and velocity.4.For multi-target detection parameters which are extracted from the above process,multiple targets’ tracks are described with the help of DSUKGMPHD algorithm so as to achieve the tracking of multiple targets.Automotive radar has been researched and made.It goes through off-site tests to checkout its capability of detecting and tracking for multiple targets.