多传感器信息融合联邦滤波一般模型的理论与仿真研究  

Theoretical and Simulation Study on General Model of Federated Filtering with Multi-sensor Information Fusion

作　　者：孙连霞[1] 穆荣军[1] 崔乃刚[1] 

机构地区：[1]哈尔滨工业大学航天工程系,哈尔滨150001

出　　处：《航空兵器》2006年第4期7-10,共4页Aero Weaponry

基　　金：国家863-702(2004AA721074)资助项目

摘　　要：对多传感器信息融合联邦滤波一般理论模型及其实现方法进行了研究。结合应用需求分析了联邦滤波一般模型的特点,给出了实用的信息分配系数求取方法及该模型具体的算法实现途径。以惯性/卫星(GPS/北斗双星)/多普勒/星光(INS/GPS/RDSS/Doppler/CNS)飞行器组合导航系统为例进行了联邦滤波一般模型算法的仿真分析。仿真结果表明,所给出的一般模型实现算法,在保持该模型全局最优性的同时,具有数值计算稳定性好、计算量小、数据传输量小等优点,是一种改进型算法。The general theoretical model and the way to put it into effect of federated filtering for multi-sensor information fusion are researched. For request of application, characteristics of general model are analyzed, the practical algorithm to get the info-partition coefficient and the specific method to realize the model are presented. Based on INS/GPS/RDSS/Doppler/CNS, a specialized integrated navigation system, math-simulating and analysis are proceeded. The results of simulation show that the way to realize general model which remain the global optimality, stability, high computational efficiency, is improved.

关 键 词：多传感器 信息融合 联邦滤波 一般模型 

分 类 号：TJ765[兵器科学与技术—武器系统与运用工程]