Improving End-to-End 6-DOF Trajectory Optimization for High-Mass Mars Landings

In this work, we test the efficacy of new improvements to the Sequential Convex Programming approach for a challenging multi-phase, multi-constrained, 6-DoF end-to-end trajectory optimization problem for entry, descent and landing of a high-mass mission on Mars. We focus on the use of line-search back-tracking methods, extended herein to a multiphase setup, in combination with a comparison of several trust-region (both soft and hard) strategies to quantify if and to what extent they affect the results. Numerical results suggest that, for the class of problems treated in this work, line-search methods show limited effectiveness, while the use of soft trust regions robustifies the trajectory optimization process. This work also helps crystallize some of the standing challenges in applying sequential convex optimization to complex, nonlinear and non-convex problems.