改进SVC算法在雷达信号分选中的应用

为提高支持向量聚类（SVC）对分布复杂、不均匀雷达辐射源信号样本分选的正确率,提出一种改进的支持向量聚类分选方法,先采用支持向量聚类对所有未知样本作预分类,提供初始的聚类中心,然后利用K-Means聚类分选算法最终分选。结果表明,此方法能够很好地对复杂雷达信号进行分选,分选正确率较高。

Application of Improved SVC Algorithm to Radar Signal Sorting

In order to increase the correct rate of support vector clustering （SVC） sorting radar radiation source signal samples with complicated and non-uniform distribution, this paper brings forward an improved SVC sorting method, which firstly uses SVC to presort all unknown samples, provides the original clustering center,then uses K-Means clustering sorting algorithm to perform the final sorting. The results show that this method can sort the complicated radar signals perfectly and the correct rate of sorting is satisfying.