摘要
使用广义总体最小二乘(GTLS,generalized total least squares)方法对零动量卫星进行惯量矩阵在轨辨识.提出了GTLS算法的先验最小距离解的定义:当测量信息不足以确定唯一解时,解空间中最接近先验估计的解.给出了先验最小距离解的算法,并应用于惯量矩阵在轨辨识.仿真结果表明了该辨识方法的有效性及先验最小距离解相对于最小范数解的优越性.
A generalized total least squares method is adopted to in-orbit identification of the inertial matrix of a zero momentum satellite. A prior minimum distance solution is defined ： this solution is closest to the prior estimate in solution space if measurement information is not enough to determine a unique solution. An algorithm of the prior minimum distance solution is proposed, and applied to in-orbit identification of an inertial matrix. Simulation results validate feasibility of the identification method and advantage of the prior minimum distance solution over the minimum norm solution.
出处
《空间控制技术与应用》
2009年第5期26-30,64,共6页
Aerospace Control and Application
关键词
零动量卫星
惯量矩阵
在轨辨识
广义总体最小二乘
先验估计
zero momentum satellite
inertial matrix
in-orbit identification
generalized total least squares
prior estimation
作者简介
林佳伟（1982-），男，福建人，博士研究生，研究方向为航天器在轨辨识（e—mail：linjw0207@gmail．com）．
