University of Colorado, Boulder
Future missions to Mars that will land large payload masses to a variety of locations across the Martian surface will require the use of propulsive control during descent. This necessitates the development of new guidance algorithms since current algorithms are for aerodynamic control only, or apply to propulsive control during the final stages of descent. This project seeks to develop these new guidance algorithms, with a focus on creating algorithms that are robust to the major sources of uncertainty during Entry, Descent, and Landing (EDL). The algorithms will incorporate machine learning methods to predict the effects of un-modeled perturbing accelerations, such as wind gusts and atmospheric conditions, and will make control decisions based on the non-linear propagation of system uncertainty to the targeted conditions.