基于Bayes理论的星间链路自主定轨误差评估方法

针对导航星座自主运行期间因缺乏外部基准而导致自主定轨结果不能得到有效评估的问题，提出一种基于Bayes理论的自主定轨结果评估方法。一方面充分考虑定轨误差的先验统计特性，以提供先验位置基准；另一方面以卡尔曼滤波中定轨协方差的迹作为定轨URE的表征量，解决了自主定轨中对定轨误差没有观测信息的问题。利用自主定轨的仿真数据对该方法进行验证，结果表明，对于单颗卫星，在24 d内，由该方法得到的URE估计误差不超过±0.5 m，验证了该方法的可行性。

Autonomous Assessment Method of Autonomous Orbit-Determination Error Based on Bayesian Theory

Due to a lack of an external reference during days the navigation constellation runs autonomously without ground support,the results of autonomous orbit determination cannot be effectively assessed. In order to resolve this problem, this paper presents the assessment of the results of autonomous orbit determination based on Bayesian theory. The method adequately takes the prior information of orbit-determination error into consideration and uses it to provide position reference. To solve the problem that on autonomous orbit determination issues there are no direct observations of the orbit-determination error, we advise using the trace of the covariance matrix of the Kalman filter as observation information for orbit-determination error. To prove the method, a simulation of the autonomous orbit determination progress is conducted.The results of the simulation show that, for a single satellite, the error of the estimated is controlled within 0.5m, in 24 days. The feasibility of the method is confirmed by the result. Using this method can provide support for autonomous integrity monitoring of on-board autonomous operation software.