Zachary Manchester
Stanford University
ECF18 Overview Chart Manchester.pdf
Managing uncertainty is a fundamental problem for entry vehicles. Knowledge of a planet’s atmosphere, winds, and the vehicle’s position and velocity are all imperfect. As future NASA missions seek to land larger payloads with greater precision on Mars and elsewhere, effectively reasoning about uncertainty will become even more crucial. To meet this challenge, we propose a unified framework for modeling, trajectory design, and control that explicitly deals with uncertainty at every stage in the process to enhance performance and safety. Our approach harnesses new tools from optimization and robotic motion planning to compute “invariant funnels” – tubes around a nominal vehicle trajectory that bound the effects of uncertainties and disturbances like winds. Using these funnels, we can plan trajectories that are guaranteed to meet safety and accuracy requirements while taking full advantage of vehicle capabilities to enhance performance.