Speaker
Description
Tens of thousands of known small Solar System bodies have orbits passing within approximately 0.5 AU of the Mars’ orbit. Identifying objects that could be redirected onto Mars-impacting trajectories under weak continuous forcing is relevant to studies of long-term planetary engineering, including conceptual investigations of Mars terraforming. Such problems also provide a demanding test case for numerical trajectory-optimization methods.
In this study, we consider asteroids previously identified as possible to redirect to Mars under a maximum acceleration of $4 \times 10^{-10}\,\mathrm{km\,s^{-2}}$. The aim is to compare the robustness and computational performance of different optimization strategies for constructing physically valid Mars-impact trajectories.
A benchmark set of 90 asteroids was selected and divided into three difficulty classes: easy, intermediate and difficult, with 30 objects in each group. Four different optimization approaches were tested:
- a direct transcription method with piecewise-linear control,
- a spline-based continuous-thrust parametrization,
- an impulsive multiple-shooting formulation - the Sims–Flanagan scheme,
- and a Bézier-curve parametrization of the thrust profile.
The results show that reliable Mars-impact targeting is possible even for objects with large natural miss distances. The two best-performing continuous control approaches recovered validated impact trajectories for all 90 benchmark cases, including all difficult targets, whereas the impulsive and Bézier-curve formulations showed reduced robustness for the most challenging objects.
These results indicate that continuous low acceleration, accumulated over multi-year to decadal timescales, can provide sufficient control authority for asteroid redirection, while the choice of optimization architecture strongly affects the robustness of the resulting solutions.