LiAISON/CNS integrated navigation method for cislunar spacecraft based on adaptive robust filtering

The linked autonomous interplanetary satellite orbit navigation (LiAISON) method enables simultaneous navigation for multiple cislunar spacecraft with unknown states through distance measurements. However, for near-coplanar cislunar spacecraft, the navigation error in the direction normal to the orbital plane increases significantly. Moreover, this method alone requires a long convergence time, typically one orbital period, and the cost of re-convergence after error perturbation is high. To address the above issues, based on observability theory, this paper analyzes the improvement effects on navigation system observability provided by Earth and Moon vectors, and the angles between inter-spacecraft line-of-sight vectors and background stars. Building on the LiAISON method, a LiAISON/celestial navigation system (CNS) integrated navigation method is proposed, which incorporates the angles between inter-spacecraft line-of-sight vectors and background star vectors. Additionally, an improved adaptive robust cubature Kalman filter algorithm is presented. It employs chi-square tests to detect measurement anomalies and adaptively estimates the process noise covariance matrix and measurement noise covariance matrix, thereby suppressing the impact of measurement outliers on filter convergence while improving orbit determination convergence speed and accuracy. Simulation results demonstrate that the proposed method achieves high-precision robust navigation for near-coplanar cislunar spacecraft.