论文部分内容阅读
为了降低卫星姿态动力学模型误差和初始姿态误差对无陀螺卫星姿态确定的影响,提出了一种预测无迹卡尔曼滤波(UKF)算法。该算法利用预测滤波中的泰勒级数展开方法对未知的卫星姿态动力学模型误差进行估计,然后将其带入UKF中实现无陀螺卫星姿态确定。用三维偏差四元数代替四元数作为状态变量,避免了估计过程四元数单位约束性被破坏的现象。仿真结果表明,当较大的卫星姿态动力学模型误差和初始姿态误差同时存在时,该算法仍能精确地实现卫星姿态确定。
In order to reduce the impact of satellite attitude dynamics model errors and initial attitude errors on the attitude determination of gyroscopes, a predictive Unscented Kalman Filter (UKF) algorithm is proposed. The algorithm uses the Taylor series expansion method in the prediction filter to estimate the unknown satellite attitude dynamics model error, and then it is introduced into the UKF to achieve gyro-free satellite attitude determination. The quaternion of three-dimensional deviation is used as the state variable instead of the quaternion, which avoids the phenomenon that the quaternion unit constraint of the estimation process is destroyed. The simulation results show that the proposed algorithm can accurately determine the attitude of the satellite when the error of the larger satellite attitude dynamics model and the initial attitude error coexist.