| With the development of social economy and the improvement of living standard,automobile,as the main tools of transportation,has become an indispensable tool in today’s social life.Though automobile brings people much convenience,the safety problem can’t be ignored.As a kind of active security defense measure,automotive anti-collision radar has gradually become the focus of attention,and millimeter wave radar has received the favor of many countries and research institutions relying on the unique advantages of small size,light weight,low cost,full-time,high resolution,etc.Some key techniques of multi-target detection and parameter estimation algorithms for millimeter wave radar are investigated and a low-complexity signal processing module is designed and applied to the 24GHz automobile front anti-collision radar system,based on the background of automotive anti-collision radar.Firstly,the research background and significance of millimeter wave radar are expounded,and the research status at home and abroad is introduced,including the development of product level research and key techniques research,and then the principles of different systems radar are derived,which provides theoretical basis for subsequent research and simulation.Secondly,multi-target pairing and velocity ambiguity resolution algorithm of Linear Frequency Modulated Continuous Wave(LFMCW)radar is investigated.For triangular LFMCW radar,a novel multi-step pairing algorithm is proposed to overcome the drawback of current algorithm with high complexity when the number of targets is large,which uses priori knowledge to reduce pairing range.As the velocity ambiguity resolution algorithm which uses multiple pulse recurrence frequency(PRF)for chirp sequence waveform radar has the drawback of high complexity and low robustness,an improved algorithm is proposed,which calculates the possible velocity value,corresponding discrete Fourier transform(DFT)factors,and spectral amplitude value according to the ambiguous value,and then the unambiguous velocity,the spectral amplitude value of which should be the biggest,can be obtained.Thirdly,the Constant False Alarm Rate(CFAR)detecting algorithm is investigated.To reduce the effect of target shadowing and self-shadowing,we design an improved CFAR algorithm refreshing the sample used to estimate noise power adaptively.The detector adds a feedback operation during detecting,and uses the estimated noise power level to replace the value of the unit where a target has been detected.For two-dimensional CFAR detecting algorithm,velocity range of the clutter in different range cell can be calculated according to the velocity of radar,based on which the two-dimensional plane is divided to two parts,called as noise and clutter region.Different CFAR detecting algorithms can be used in different regions to improve detection probability under the premise of ensuring false alarm probability.Finally,aiming at the problem that the frame structure of the automotive anti-collision radar with variable transmitting period is not clear and practical,a multi-level and low-complexity frame structure as well as its designing procedure is proposed.The frame structure takes sampling interval as the smallest time unit,which is more conducive to the system synchronization,and fixed length of time slot and sub time slot makes the frame structure simpler and reduce the cost of hardware implementation.To achieve more beamforming gain,an equivalent virtual array with larger aperture is designed by adjusting transmitting antenna space and transmitting signal through time diversity.Based on these,a low-complexity signal processing module is designed,which consists of several sub-modules including pre-processing,beamforming,CFAR detection and multi-target pairing,etc.The signal processing module has been implemented on the ADSP-BF707 platform,and applied to the 24GHz automobile front anti-collision radar system.The measured result shows that the signal processing module can make target detection and parameter estimation. |
PDF Full Text Request