卫星多敏感器组合姿态确定系统中的信息融合方法研究

针对卫星陀螺，红外地平仪，太阳敏感器，GPS接收机组合姿态确定系统的特点，提出一种基于联邦卡尔曼滤波算法的信息融合方法，其中信息分配系数通过计算协方差矩阵的迹在线自适应确定。推导了由四元素描述的卫星姿态误差状态方程和各子系统的测量方程。仿真分析结果表明采用该信息融合算法可以提高定姿精度，有效抑制滤波发散，并使整个系统的运算速度和收敛速度都有所提高。

Research On Information Fusion Method In Satellite Multi-sensor Attitude Determination Systems

Based on federated Kalman filter, a new information fusion algorithm in the solution of state estimation problem arising from satellite multi-sensor attitude determination systems is presented. The information sharing factors are dynamically determined according to the traces of the covariance matrix. State error equation represented by quaternion error and measurement equations of local filters are given. The simulation results indicate that divergence of filter can be inhibited and the accuracy and convergence speed to the whole system can be improved.