一种改进的聚类目标融合算法

An improved algorithm of clustering target fusion

多源目标数据融合技术是雷达辐射源关联融合中的一项关键技术。以K均值算法为基础,对簇中心初始化方法进行优化,提出了基于空间密度与欧氏距离结合的聚类初始化算法,对聚类方法进行改进,并将其应用于多源目标融合领域。仿真结果验证了该算法可以有效地提高多源融合性能。

Multi-target data fusion is one of the key techniques in the association and fusion of radarradiation sources. A new initialization method for K？means clustering by utilizing spatial density andEuclidean distance is proposed, and the initialization method of clustering center is optimized andapplied in the field of multi？target data fusion. The simulation results show that this algorithm can effectively improve the multi？target data fusion performance.