摘要
针对本体坐标系下对卫星姿态进行线性滤波存在误差累积和噪声分布改变无法正确选择滤波模型的问题,利用惯性坐标系下的星敏感器和陀螺原始测量构建系统扩维测量,将卫星姿态确定问题建模为非线性滤波过程。针对常用的非线性滤波方法无法同时兼顾精度和实时性的问题,采用无迹滤波(UKF)对卫星的姿态参数和陀螺常值漂移同时进行估计,实现了对卫星精确定姿。进一步考虑实际情况中,敏感器测量误差未知或随时间变化的情况,提出了交互式多模型无迹滤波(IMMUKF)方法。仿真实验结果表明了该算法的有效性和优越性,具有较高的实际应用价值。
Aiming at the problem of unable to choose suitable filtering mode for errors accumulated and noise distribution changed,the body coordinate system,a augmented measure composed of measurements of star-sensor and gyro in the inertial coordinate is proposed and the attitude determination is modeled to a nonlinear filter problem.Because the nonlinear filters used commonly may have a good performance both in accuracy and real time,UKF is introduced to estimate the attitude parameters and the gyro drift.Further,the situation of unknown or time-varied sensor accuracy is considered,and the IMMUKF algorithm is proposed.The simulation results show the efficiency and accuracy of the method.It has a high practical value.
出处
《现代电子技术》
2012年第4期13-17,共5页
Modern Electronics Technique
关键词
姿态确定
星敏感器
陀螺仪
无迹滤波
交互式多模型
attitude determination
star-sensor
gyro
unscented Kalman filter
interacting multiple model
