Data-Driven Representations of Trajectories in Cislunar Space

Autonomous trajectory design in cislunar space poses a difficult challenge. Due to the nonlinear, chaotic nature of the system, there is no simple representation of trajectories akin to the orbital elements of two-body motion. Various structures such as periodic orbits, quasi-periodic orbits, and the stable/unstable manifolds of these orbits have properties which can benefit missions. For example, quasi-periodic orbits and their invariant manifolds can help spacecraft traverse cislunar space with low fuel expenditure. However, due to the complex analysis needed to compute these trajectories, they are often not considered for mission design and cannot be used by autonomous systems in motion planning. This project addresses this issue by using data-driven methods to model trajectories in the Circular-Restricted Three-Body Problem and eventually higher-fidelity models of cislunar space with a low-dimensional representation. This would allow autonomous systems to reference these orbits by analyzing their components in this reduced dimension and reconstruct nearby trajectories that are accessible to the spacecraft and beneficial to its mission objectives. To do this, we employ state-of-the-art techniques in machine learning that are being used in the dynamical systems community. The intersection of machine learning and dynamical systems is rapidly expanding, and two methods show exceptional promise, Koopman operators and Deep Autoencoders. Both tools can represent large datasets of nonlinear trajectories in a low-dimensional space and easily reconstruct the full trajectories. As NASA's Gateway and Artemis programs increase our presence in the cislunar environment, autonomous motion planning is an important focus of the NASA Strategic Framework. Our project would allow for pretrained models of cislunar trajectories to be loaded into onboard computers, reducing the cost associated with developing trajectories and expanding the types of trajectories and operations accessible to autonomous spacecraft.

This is a mature technology (TRL 7+) — the realistic path in is usually NASA's Technology Transfer Program: licensing an existing NASA patent, or a Space Act Agreement to use NASA facilities/expertise directly. NASA also runs a startup licensing program with no upfront fee for companies formed to commercialize a specific NASA technology.

None of these are guaranteed paths for this specific project — TechPort itself doesn't have an "apply" button. Reaching out to the contact(s) above with a specific question is usually the fastest way to find out what's actually open.