Formation Satellites Positioning Based on Cubature Kalman Filtering and Factor Graph Optimization

High-precision satellite positioning is vital in modern aerospace systems, especially for satellites formation systems with requirements of sub-meter-level control. This paper presents a cascaded ranging information fusion framework based on a cubature Kalman filter (CKF) and factor graph optimization (FGO) for formation satellites positioning. The GNSS ranging information and inter-satellite ranging information are used. Considering the improvement of the state estimation performance by the Kalman filter in satellite positioning systems and the sensitivity of FGO to initial values, a CKF is utilized to filtering GNSS ranging information, achieving sub-meter-level initial satellite position estimates. Then, the estimation results from the CKF are used as the initial values for the FGO iteration. By integrating GNSS ranging factors and inter-satellite ranging factors, the precise positioning of the satellite formation can be obtained. Through simulation experiments conducted on a simulated satellite formation positioning system designed in STK software environment, it is demonstrated that the presented CKF-FGO-based multi-source information fusion algorithm can significantly improve the formation satellites positioning accuracy.