| The radial Doppler frequency of the target relative to radar motion is small,Therefore,traditional methods such as comparing the beat frequency of sweep signals or calculating the displacement difference of targets and the commonly used range-Doppler processing can not be used to directly estimate the velocity of tangential moving targets,and the existing traditional moving target detection methods can not effectively detect tangential moving targets.In order to overcome the above problems,this paper proposes a tangential target velocity measurement method which uses multi-channel echo data to measure the moving target perpendicular to the radar illumination beam.At the same time,a step-by-step CFAR-GLRT detector is used to detect targets in different stages,aiming at the suboptimal Bayesian decision method that only uses the traditional CFAR detection method to select the threshold value according to the background clutter without considering the statistical characteristics of the target,and usually uses the sliding window method with a large amount of computation.Because the traditional coherent integration algorithm only uses the received signal of a single radar and the signal in a single coherent period for target detection,and does not use all the information of the received signal,Therefore,the non-coherent accumulation algorithm based on multi-radar and multi-coherent period is considered,which uses the signals of multi-radar and multi-coherent period to detect the target and combines the particle filter method to realize the effective detection of weak target.In addition,a clutter map algorithm based on deconvolution is proposed to eliminate clutter iteratively,combining the shortcomings of traditional spectrum thinning and clutter map algorithm.Finally,the algorithm is tested on the IWR1443 radar development board and DCA1000 data acquisition board of Texas Instruments,which proves the feasibility of the algorithm.Figure 55;Table 4;Reference 41... |
PDF Full Text Request