因子图在卫星姿态估计中的应用  

Estimation of Satellite Attitude Angular Velocity Based on Factor Graph

作　　者：陈恩庆[1] 肖素珍[1] 高新利[1] 

机构地区：[1]郑州大学信息工程学院,河南郑州450001

出　　处：《计算机仿真》2015年第6期63-66,共4页Computer Simulation

基　　金：国家自然科学基金(61201251;61172086;61210005;61331021)

摘　　要：在北斗卫星导航系统定位过程中首先要对卫星自身的姿态角速度进行估计。卫星姿态角速度估计的精度将直接影响北斗卫星定位的精度。为了提高北斗卫星姿态角速度估计的精度,提出了一种新的采用概率图模型的卫星姿态角速度估计方法,即基于因子图消息传递的卫星姿态角速度估计方法。因子图是一种典型的概率图模型方法,它通过对卫星姿态角速度观测方程和状态方程建模,并使用和积迭代算法求解相关参数,获得对卫星姿态角速度的估计。通过MATLAB仿真,对所提算法与传统基于kalman滤波的姿态角速度估计算法进行比较,结果表明改进算法可以准确估计卫星的姿态角速度,并且在低信噪比情况下比kalman滤波算法有更小的估计误差,算法收敛速度更快,可为北斗卫星导航定位系统优化设计提供参考。In order to improve the estimation accuracy of beidou satellite attitude angular velocity, this paper pro- poses a new estimation method of satellite attitude angular velocity based on factor graph. Factor graph is a kind of typical probability graph model, the satellite attitude angular velocity is modeled by an observation equation and a state equation, the related parameters are obtained by using the iterative algorithm, and the estimate of the satellite attitude angular velocity is got. Through MATLAB simulation, the proposed algorithm is compared with the traditional attitude angular velocity estimation algorithm based on kalman filter, and the results show that the algorithm can ac- curately estimate the satellite attitude angular velocity, and under the condition of low signal - to - noise ratio, the estimate error is smaller than the kalman filter algorithm, in addition, the algorithm has faster convergence speed.

关 键 词：姿态角速度估计 因子图 消息传递算法 卡尔曼滤波 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]