基于自适应算法的月球引力常数和J_2系数确定

Autonomous Determination of Lunar Gravity Constant and Coefficient of the Lunar J_2 Perturbation Based on Adaptive Kalman Filter

针对月球引力场自主确定过程中测量噪声统计特性未知导致扩展卡尔曼滤波精度低、易发散的问题,提出了一种自适应扩展卡尔曼滤波算法。该算法通过采用改进的噪声估计器,对滤波过程中未知测量噪声统计特性进行实时估计和修正,有效地提高了扩展卡尔曼滤波器的稳定性,减小了状态估计的误差。通过与蒙特卡洛仿真,扩展卡尔曼滤波的结果比较,自适应扩展卡尔曼滤波算法加强了滤波的稳定性,并且明显提高了月球探测器轨道的确定精度、月球引力常数精度和月球J2项摄动系数精度。

In terms of the low filter accuracy and divergence caused by unknown measurement noise statistics in autonomous determination of the lunar gravitational field,an adaptive extend Kalman filter（ AEKF） is presented. By adopting the modified noise statistic estimator and proposing a new method to restrain the divergence of the traditional extend Kalman filter,the state estimation accuracy is effectively improved and the filtering divergence is valid controlled.Compared with the results from Monte Carlo simulation and extend Kalman filter,the adaptive extend Kalman filter not only enhances the stability of the filter,but also reduces the state estimation error of the determination of the lunar probe＇s orbit,it can also achieve the high accuracies of the lunar gravity constant and the coefficient of the lunar J2 perturbation.