A Stereo-Vision Based System for Autonomous Navigation of an In-Orbit Servicing Platform

This paper deals with the autonomous relative navigation of two satellites flying in close proximity. The subject is studied in the framework of a national project financed by the Italian Space Agency to test in space advanced technologies for formation flying and in-orbit servicing missions. Specifically, a small platform will be realized embarking a set of sensors and actuators to safely fly around a larger satellite to carry out a set of experiments ranging from 3D satellite monitoring to autonomous navigation and control. This paper focuses on the proposed navigation algorithm which consists of a real-time Kalman Filter integrating a linearized model of the relative orbital dynamics with the measurement of a stereo-vision system. Navigation system accuracy, effectiveness and robustness are numerically tested considering realistic orbital scenarios.