多传感器数据分层融合的性质

本文主要研究文［１］提出的分层融合算法的数学、物理性质。定义了分层融合估计中的融合航迹的新息和融合增益，分析并证明了它们所具有的数学、物理性质，最后得出并证明该分层融合估计，就是基于各传感器的滤波及其预测的线性无偏最小方差估计的结论。

The Properties for Hierarchical Fusion of Multisensor Data

The mathematical and physical properties of an algorithm for hierarchical fusion proposed in 1] are studied in this paper. First, the concepts for innovation and gain of fusing tracks are proposed. Second, their mathematical and physical properties are analyzed and proved.Final, we show that the hierarchical fusion estimate is a linear, unbiased, minimum variance estimate based on filter and prediction of each sensor. These properties reveal the essence of the hierarchical fusion eseianate in 1] and have of great value for solving problem on multisensor multitarget.