LABORATORY EXPERIMENTS SUPPORTING AUTONOMOUS SPACE DEBRIS MITIGATION
Additional Document Info
High fidelity ground-based simulation is a necessary step in the development of successful autonomous orbital debris removal missions. A guidance, navigation, and control package is presented, which supports such experiments. A structured light stereo sensor delivers three-dimensional point cloud data to an object recognition module. The resulting navigation solution comprises the relative position and attitude of a tumbling target vehicle with respect to the capture vehicle. An onboard filter consumes these relative vision measurements along with data from an inertial measurement unit, providing a mission handler with the best estimate of relative position and attitude. Built on an event-driven framework, the controller maximizes target visibility to the vision sensor and drives a probe down the throat of the target nozzle. A dynamic vehicle emulator translates the motion of a virtual spacecraft of choice into real motion of the robotic experimental platform. This paper presents the object recognition algorithm, filter and control components, and calibration procedures used in the development of the GNC package along with experimental results for two mission scenarios. Copyright 2013 by The Math Works, Inc.