Autonomous navigation for lunar satellite using X-ray pulsars with measurement faults

摘要

X-ray pulsar-based navigation (XNAV) is a navigation method using celestial X-ray source observations for spacecraft orbit determination. However, if the measurements are not reliable due to any kind of malfunction, the performance of XNAV may result in considerable errors and even divergence. In this study, a new algorithm called robust extended Kalman filter (REKF) is proposed for the lunar satellite autonomous navigation system, which is robust against measurement malfunctions. First, the satellite dynamic model applied perturbations is derived under the J2 perturbation of the Moon. The pulse time-of-arrival (TOA) is used to build the observation model. Second, the performance of XNAV system is discussed by analysing the transformation error of TOA, system observability and controllability. Then, an adaptive measurement noise scale factor (AMNSF) is designed by using the innovation sequence in REKF. Meanwhile, the gain matrix is modified by adding the AMNSF to reduce the influence of malfunction and enhance the robustness of XNAV system. Finally, the simulation results show the proposed navigation scheme is valid and feasible, and it is appropriate for lunar satellite autonomous navigation.