What is it about?
Reinforcement learning has been proven to be powerful. How can we utilize it in space missions, such as building structures in space autonomously? In our paper, we have shown that we can employ it as a search agent to find the optimal trajectory, in terms of fuel consumption, for a spacecraft assembling a piece of structure to another in deep space.
Featured Image
Photo by Marcus Urbenz on Unsplash
Why is it important?
As humans go deeper into space as part of their space exploration goal, they need to build large, complex structures at deeper locations in space, such as research facilities, habitats, refueling stations, commanding units, etc. Due to long distances and the resultant communication challenges, as well as the size and mass limitations of launch vehicles, the construction of such structures in deep space should take place by autonomous robotic spacecraft rather than astronauts. This would not be possible unless appropriate algorithms, such as our work in this paper, were developed.
Perspectives
In my eyes, reinforcement learning is a powerful tool that can be employed in many applications. I think that more and more challenging problems in aerospace engineering can be addressed more conveniently using reinforcement learning. My co-authors and I hoped to demonstrate that this powerful tool can be creatively applied to various issues in aerospace engineering.
Siavash Tavana
Ryerson University
Read the Original
This page is a summary of: A Reinforcement Learning-Based Continuation Strategy for Autonomous On-Orbit Assembly, January 2024, American Institute of Aeronautics and Astronautics (AIAA),
DOI: 10.2514/6.2024-0958.
You can read the full text:
Contributors
The following have contributed to this page
