在时域信号中引入FCM算法,以区分运动目标的多普勒频移特性.利用多普勒特性对雷达中频信号进行了模拟研究,并使用FCM算法在一个雷达扫描周期可确定两个聚类中心,分别对应两个静止和运动的目标信号;对多个雷达扫描周期的多个聚类中心点,采用算术平均值滤波的方法构造中心函数C,通过门限判决的方法判定目标,并计算目标的检测概率.该算法简单、易于实现、计算量小,且有良好的目标检测概率.%In this paper, Doppler shift was applied to simulate a radar medium frequency signal. Since it is difficult to distinguish the Doppler shift of moving target even in frequency domain with FT and STFF method, fuzzy C means clustering method was introduced to seek the two centers of the mixed signals in a radar scanning period. As for multiple radar scanning periods, several centers were determined by a center function using arithmetic mean filter. Finally the targets were detected by threshold decision and the detection probability. This algorithm is simple, easy to realize, with small computation burden and good target detection probability.
展开▼
