Published: Aug. 30, 2021

Natasha BosanacNatasha Bosanac
Assistant Professor, Smead Aerospace
Wednesday, Sept. 8 | 12:00 P.M. | Zoom

Abstract: Spacecraft technology is undergoing a significant advance that will expand humanity’s presence in space. Our future involves miniaturized satellites for low-cost and rapid access to space, on-orbit servicing to increase the lifetimes of spacecraft, in-space assembly of critical infrastructure, formations for multi-point measurements, and spacecraft visiting the farthest reaches of our solar system.

Across these architectures, form factors, and destinations is a common thread: spacecraft operating in complex, multi-body systems under the influence of multiple gravitational bodies. In these regimes, trajectory and maneuver design strategies currently rely heavily on a human expert who can perform complex analyses, identify patterns, make decisions and adapt as needed.

However, advancements in trajectory and maneuver planning that move away from this paradigm can enhance missions by reducing operational cost and complexity or enable new missions by, for example, supporting rapid response to uncertain environments without reliance on a ground-based team. Thus, trajectory and maneuver design emerges as a critical enabling and enhancing technology.

In this talk, I will present recent contributions from our research group that focus on using techniques from reinforcement learning, unsupervised learning and supervised learning to reduce the burden and reliance on a human-in-the-loop during keys steps in trajectory and maneuver design within multi-body systems. I will also highlight our recent contributions that focus on using trajectory design to enable new types of missions.

Bio: Natasha Bosanac is an Assistant Professor in the Colorado Center for Astrodynamics Research within the Smead Aerospace Engineering Sciences department at the University of Colorado Boulder. She specializes in applications of dynamical systems theory to astrodynamics in multi-body systems. Her research group is currently working on advancing trajectory and maneuver design strategies through a combination of dynamical systems theory, machine learning and path planning techniques.

 

Upcoming Seminars Calendar

All seminars are open to the public.