一种分布式信息融合的航天器实时轨道确定算法

A Distributed Information Fusion Based Algorithm for Spacecraft Real-Time Orbit Determination

针对传统航天器集中式实时轨道确定方法对计算中心依赖性强、鲁棒性差等问题,在分布式混合信息滤波算法(DHIF)的基础上,提出一种基于分布式信息融合的航天器实时轨道确定算法。通过该算法各测站融合自己和邻居节点的状态信息及量测信息,可同时对目标航天器进行局部实时定轨。该算法支持包含多类型传感器的观测网络,对定轨系统局部变化适应性强。仿真结果表明,该分布式定轨算法使各测站局部定轨精度优于单站滤波定轨精度,且一致逼近于多站集中式滤波定轨精度,测站局部定轨的收敛速度取决于测站网络通信的拓扑结构。

Aiming at the problems which exist in the centralized real-time orbit determination method,such as strong dependence on computing center and poor robustness,this paper proposes a distributed real-time orbit determination algorithm on the basis of distributed hybrid information filtering(DHIF).By fusing the state information and measurement information of each station and its neighbors,the local real-time orbit determination can be carried out simultaneously.This algorithm can be applied to an observation network with multi-type sensors,and is adaptive to the locally unknown changes in the network.The simulation results indicate that the accuracy of the distributed orbit determination algorithm is better than that of single station information filtering,and the estimate of each station converges to a centralized estimate.The convergence speed of the local orbit determination depends on the topology of the communications.