Vision Navigation Performance for Autonomous Orbital Rendezvous and Docking

This thesis demonstrates the potential of performing orbital rendezvous and docking using vision navigation. The vision navigation algorithm tracks both known and unknown target features to determine the relative position and attitude between a chaser and target spacecraft. By processing imagery generated from an optical sensor, various target features can be tracked to accurately determine the relative motion between two orbiting vehicles. This research adopts an architecture that uses an extended Kalman filter (EKF) to processes angle measurements to various target features as extracted from the vision navigation algorithm. One potential limitation to this approach is determining the image scale or range. A Monte Carlo simulation evaluates the performance of the navigation filter in a closed-loop guidance, navigation, and control (GNC) system. This research introduces strategies to overcome the resulting range dilemma and characterizes the performance of using vision navigation for autonomous orbital rendezvous and docking.