Vortragsdetails

PhD Colloquium Marko Jankovic: Characterization and Capture of Space Debris Objects using Domain Ontology and optimal Control

An ever-increasing number of in-orbit objects coupled with inadequate post-mission mitigation activities in certain orbital regions is causing a reaction from the space community more then ever. On one hand, a more sustainable usage of orbital environment is being explored while on the other, methods to actively reduce the in-orbit mass are being researched. Selecting one method over another, even for a specific object, is nonetheless not an easy task. This is mainly due to the amount of information associated with each object. At the same time, the lack of a common, standardized, machine-interpretable framework able to store that information does not facilitate matters. Among the currently investigated active debris removal (ADR) methods, robotic-based systems are perceived as the most mature and versatile ones.
But they introduce challenges of their own especially in case of uncooperative objects characterized by high angular momenta (i. e. high angular rates and/or inertia). The reason for this can be found in the strict safety requirements associated with such an ADR method and its ability to manage the angular momentum post-capture. Using a domain-ontology, specifically developed for ADR, the first objective of the thesis is to provide a standardized, machine-interpretable knowledge representation framework. This framework is capable not only of efficiently storing complex information but also inferring the most suited ADR capture method(s) for an object from its minimal set of parameters. The second objective of the thesis is to provide an optimization-based motion planner of a spacecraft equipped with a manipulator, a space robot. The purpose of the planner is to facilitate an autonomous capture of an uncooperative, rapidly tumbling target by means of a space robot. This is achieved by taking the advantage of the  particular dynamics of the space robot and an appropriate partial attitude synchronization strategy. By doing so the planner is able to find an optimal coordinated motion of the whole spacecraft to approach a rapidly tumbling object while minimizing its overall control effort. The practicality and validity of the ontology are demonstrated by applying it onto a database of representative objects.
The latter was built by combining structured and unstructured data from publicly available sources. The analysis of results proves the ontology capable of inferring the most suited ADR capture methods for considered objects. Furthermore, it confirms its ability to handle the input data from different sources transparently, minimizing user input.
This should make the initial planning of future ADR missions simpler, yet more systematic. The performance of the planner is evaluated in 3D simulation environment using different test cases. Each test case is characterized by different kinematic configuration of the manipulator and angular velocities of the object. The analysis of results confirms the ability of the planner to provide optimal, well-behaved state trajectories of the spacecraft under a range of different conditions. In this manner, a future application of the planner should facilitate close-proximity operations to objects with high angular momenta using a space robot.

In der Regel sind die Vorträge Teil von Lehrveranstaltungsreihen der Universität Bremen und nicht frei zugänglich. Bei Interesse wird um Rücksprache mit dem Sekretariat unter sek-ric(at)dfki.de gebeten.

© DFKI GmbH
zuletzt geändert am 30.07.2019
nach oben