Rapid AI-Powered Trajectory Optimization for Resupply

ATA Engineering, Inc., (ATA) proposes a Phase II Small Business Innovation Research program to develop a framework for rapid end-to-end trajectory optimization for point-to-point rocket cargo missions. When the changing battlespace requires agile adjustment to a nominal rocket cargo mission design, the proposed Rapid AI-Powered Trajectory Optimization for Resupply (RAPTOR) software framework will couple well-established tools from the trajectory propagation and optimization community with novel machine learning methods to enable rapid identification of new trajectories. Optimization of the end-to-end trajectory, including launch and ascent, on-orbit, and reentry phases, will be demonstrated using a hybrid approach combining traditional optimization and machine learning methods to enable rapid convergence to feasible trajectories. RAPTOR will include consideration of a variety of phenomena relevant to the end-to-end rocket cargo mission. Several of these phenomena, such as the reentry aerothermal environment and resulting vehicle thermal protection system response, are prohibitively computationally expensive to evaluate within an optimization framework using state-of-the-art physics-based models; RAPTOR will deploy novel reduced-order modeling methods to overcome this challenge and ensure that the generated optimized trajectories are survivable. Finally, uncertainty will be propagated throughout the entire RAPTOR framework and used to quantify trajectory dispersion and frequency of any constraint violations, as well as to support user assessments of reduced-order model trustworthiness and risk-based selection of optimized trajectories.