Application of Reinforcement Learning: Trajectories Resilient to Missed Thrust Events (MTEs), Year 2

Status: Completed

Start Date: 2020-10-01

End Date: 2021-09-30

Description:

Advance the state of the art in low-thrust mission design by developing innovative, artificial intelligence-based (reinforcement learning) techniques for designing trajectories resilient to missed thrust events (MTEs). Develop a prototype tool employing the advanced algorithms. Demonstrate increased resilience, coverage, and reduction in analysis time as compared to current techniques.

Benefits:

Develop methods for designing low thrust trajectories that are resilient to trajectory uncertainties and anomalies such as temporary thrust outages. Mission trajectory design is currently a deterministic process. Methods for dealing with uncertainties are coarse and labor intensive.

Lead Organization: Jet Propulsion Laboratory