摘要
为解决车载毫米波雷达目标检测过程中在相邻多目标情况下会提高检测门限值,从而导致目标遮蔽现象,引发传统恒虚警(constant false alarm rate,CFAR)检测在多目标环境下产生漏检的问题,提出一种等分剔除恒虚警检测算法(equipartition deletion CFAR,ED-CFAR).通过将目标两侧的参考单元等分取平均值得到子参考单元,对大于所有参考单元平均值的部分进行处理后排序,剔除部分较大子参考单元,计算得到门限值,与目标的信号幅度进行比较判断目标是否存在.仿真结果表明,在多目标检测的仿真场景下,信噪比为 10 dB时,ED-CFAR算法的检测概率比单元平均恒虚警(CA-CFAR)算法提高 0.39,比筛选平均恒虚警(CMLD-CFAR)算法提高了0.08,有效解决了传统均值类恒虚警检测在相邻多目标条件下的遮蔽问题.
Abstract
In order to solve the problem of raising the detection threshold in the condition of adja-cent multiple targets,which leads to the missed detection in the multi-target environment,amillime-ter wave radar CFAR algorithm based on equipartition deletion is proposed.The reference units on both sides of the target are equally divided into average values to obtain sub-reference units.Parts larger than the average of all reference units are sorted after processing.Some larger sub-reference units are eliminated.The threshold is calculated and compared with the signal amplitude of the tar-get to determine whether the target exists.When the SNR is 10 dB,simulation results show that de-tection probability of ED-CFAR is 0.39 higher than that of cell averaging CFAR(CA-CFAR)and 0.08 higher than that of censored mean level detector(CMLD-CFAR).The target masking problem of mean level CFAR is effectively solved in the case of adjacent multiple targets.
NETL
NSTL
万方数据