多传感器融合导航方法的比较  

Comparison of Several Integrated Navigation Methods by Using Multiple Sensor Information

作　　者：杨元喜[1] 

机构地区：[1]西安测绘研究所

出　　处：《测绘科学与工程》2006年第2期1-3,10,共4页Geomatics Science and Engineering

基　　金：国家杰出青年基金项目（49825107）和自然科学基金项目（40474001和40274002）资助.

摘　　要：多源传感器融合导航有多种算法，但不同算法之间存在理论和计算方面的差异。本文试图从理论上探讨联邦滤波融合法、动静态滤波融合法、基于观测信息的融合法以及基于各传感器几何导航解融合法的基本原理，分析各种方法的优劣，为多传感器融合导航的应用打下基础。There are several different integrated navigation methods for the multiple sensor navigation systern. There exist some differences in theory and in calculation efficiency between the various data fusion methods, This paper discusses the basic principles, calculation efficiency, fault tolerant ability of the federated Kalman filtering, synthetic Kalman filtering composed by kinematic filtering and static filtering, integrated filtering on the basis of mufti sensor observations, and integrated filtering on the basis of local geometrical navigation outputs. The research conclusions are helpful for the practical applications of the multi sensor navigation system.

关 键 词：多传感器导航 融合导航 KALMAN滤波 联邦滤波 

分 类 号：P228[天文地球—大地测量学与测量工程]