Autonomous navigation is one of the main enabling technologies for future space missions. While conventional spacecraft are navigated through ground stations, their employment for deep-space CubeSats yields costs comparable to those of the platform. This paper introduces an extended Kalman filter formulation for spacecraft navigation exploiting the line-of-sight observations of visible Solar System objects to infer the spacecraft state. The line-of-sight error budget builds upon typical performances of deep-space CubeSats and includes uncertainties deriving from the platform attitude, the image processing, and the performances of the sensors. The errors due to the low-thrust propagation and light-time delays to the navigation beacons are also taken into account. Preliminary results show the feasibility of the deep-space autonomous navigation exploiting the line-of-sight directions to visible beacons with a 3-sigma accuracy of 1000km for the position components and 2 m/s for the velocity components.
Georgia Institute of Technology（佐治亚理工学院）；Purdue University（普渡大学）；University of South Florida（南佛罗里达大学）；Skidaway Institute of Oceanography, University of Georgia（佐治亚大学斯基德韦海洋研究所）；University of Georgia（佐治亚大学）
Learning Robust Autonomous Navigation and Locomotion for Wheeled-Legged Robots
Robotic Systems Lab, ETH Zurich（苏黎世联邦理工学院机器人系统实验室）；Neuromeka；Swiss-Mile Robotics AG（Swiss-Mile机器人公司）；The work was carried out during their stay at（该工作是在他们停留期间完成的）
Integrated YOLOP Perception and Lyapunov-based Control for Autonomous Mobile Robot Navigation on Track
Graduate School of Science and Technology（科学与技术研究生学院）；Master’s Program in Science and Technology（科学技术硕士项目）；Green Science and Engineering Division（绿色科学与工程系）
