用于探测日地空间磁环境的磁强计多数安装在伸杆的末端,长期受太阳辐射等空间环境干扰力矩以及机动等影响,磁强计安装矩阵随时间发生较大的变化,从而导致卫星定姿精度下降。为此,在分析空间环境干扰力矩和磁强计定姿误差特性的基础上,建立了19维高精度的磁强计误差模型,结合卫星的运动学和姿态动力学特性,采用EKF滤波方法对安装矩阵进行实时估计与修正补偿,并利用该磁强计模型实现卫星的姿态确定,最后利用实验进行验证。实验结果表明,该方法能够在满足星载计算机的计算量要求的同时,在线估计安装矩阵误差,显著提高了磁强计的误差估计精度与定姿精度。
A magnetometer used in the sun-earth space exploration satellite is usually assembled at the tip of the boom by the action of space disturbance torque and maneuver, the installation matrix magnetometer of varies dramatically, which may lead to the low attitude determination accuracy. Based on the analysis of the magnetometer attitude determination errors, a 19 state high-fidelity measurement model of magnetometer is proposed. Combined with satellite attitude dynamics and kinematics, a 19 state Extended Kalman Filter is adopted to estimate installation matrix on-board and compensate magnetometer measurement. Then the innovative magnetometer measurement model is used to estimate satellite attitude. Finally the algorithm is validated by using the turntable experiment. Results of turntable experiment show that this method can estimate installation matrix errors under the computational requirement of On-board Computer and dramatically improve the accuracy of attitude determination and magnetometer error estimation.