【摘要】 随着辅助驾驶技术对高精度的不断追求，毫米波雷达在目标检测领域的应用逐渐成为研究热点。针对毫米波雷达在对多目标检测时门限被抬高而产生误差的情况，提出了一种融合标准差和距离的恒虚警检测算法（Standard Deviation Smallest of CFAR,SDSO-CFAR），通过对目标两侧较小的参考单元加入标准差和距离作为动态阈值，使得波动特征更大的信号门限值更低，显著提升了多目标检测的准确性并降低目标检测对信噪比的依赖。通过数据仿真验证，所采用的目标检测算法在新型雷达芯片配置下，在信噪比为10 d B时，SDSO-CFAR算法的检测概率比CA-CFAR算法高0.08，比SO-CFAR算法高0.23，有效解决了多目标检测存在的遮蔽问题。更多还原

【Abstract】 This paper addresses the issue of threshold elevation and resulting errors when millimeter-wave radar detects multiple targets, and proposes a constant false alarm rate detection algorithm that integrates standard deviation and distance. By incorporating standard deviation and distance into the smaller reference cells on both sides of the target as dynamic thresholds, the threshold value for signals with greater fluctuations is lowered, significantly enhancing the accuracy of multi-target detection and reducing the dependence on signal-to-noise ratio. Through data simulation verification, the adopted target detection algorithm, under the configuration of new radar chips, shows that at a signal-to-noise ratio of 10 d B,the detection probability of the SDSO-CFAR algorithm is 0.08 higher than that of the CA-CFAR algorithm and 0.23 higher than that of the SO-CFAR algorithm.更多还原