一种基于网格划分数据场的雷达信号分选方法

针对传统聚类算法需要人工预先设定聚类数目并且对聚类中心的选取十分敏感等问题，提出了一种网格划分数据场的雷达信号分选方法。首先，使用网格进行子空间的划分，根据网格长度确定数据场影响因子，并清除信号交叠处的高密度网格；然后，利用改进的等效距离计算数据场以确定聚类数目和聚类中心；最后，使用K均值聚类(K means)算法对雷达信号进行分选。仿真结果表明，该方法能够提高数据场的计算速度，较好地确定交叠雷达信号的聚类数目和聚类中心，并且能有效分选出参数捷变的雷达信号。

A radar signals sorting method based on grid division data field

For the problem that traditional clustering algorithms need to manually pre set the number of clusters and are very sensitive to the selection of cluster centers,a radar signal sorting method based on grid division data field is proposed.This method first uses the grid to divide the subspace,determines the data field influence factor according to the grid length,and removes the high density grid at the signal overlap,then uses the improved equivalent distance to calculate the data field to determine the number of clusters and the clustering centers,and finally uses the K means clustering algorithm to sort the radar signals.The simulation results show that this method can improve the calculation speed of the data field,determine the number of clusters and cluster centers of overlapping radar signals well,and can also effectively sort out radar signals with agile parameters.