Shaping Asteroids with Genetic Evolution (SAGE)

Abstract

In this work we present SAGE (Shaping Asteroid models using Genetic Evolution) asteroid modelling algorithm based solely on photometric lightcurve data. It produces non-convex shapes, rotation axes orientati and rotational periods of asteroids. The main concept behind a genetic evolution algorithm is to produce random populations of shapes and spin orientations by mutating a seed shape and iterating the process until it converges to a stable global minimum. To test SAGE we have performed tes on five artificial shapes. We have also modelled (433) Eros and (9) Meti asteroids, as ground truth observations for them exist, allowing us to validate the models. We have compared derived Eros shape with NEAR Shoem model and Metis shape with adaptive optics and stellar occultation observations as with other available Metis models from various inversion methods.