Automated Asteroid Selection For A ‘Grand Tour’ Mission
- Paper number
IAC-07-C1.7.07
- Author
Dr. Dario Izzo, European Space Agency (ESA)/ESTEC, The Netherlands
- Coauthor
Dr. Tamas Vinko, Hungary
- Coauthor
Dr. Claudio Bombardelli, Advanced Concepts Team, The Netherlands
- Coauthor
Mr. Stefan Brendelberger, Germany
- Coauthor
Mr. Simone Centuori, Italy
- Year
2007
- Abstract
Missions to asteroids are currently at the centre of the attention of national space agencies, academic communities and media. Deep Impact, Hayabusa, Dawn are just famous and successful names among a larger number of ongoing efforts aimed at improving our understanding of the potential benefits and threats asteroids offer to human kind. A number of missions aimed at obtaining the largest possible set of information on asteroids are being studied at different detail levels, from conceptual (e.g. SWARM and ANTS) to phase-A design (e.g. Don-Quijote) to operations (e.g. Dawn).
From a mission design point of view, asteroid mission designs are strongly driven by the initial asteroid choice, often constrained by scientific rationales, but otherwise left free to the mission designer. Jet Propulsion Laboratories (JPL) has given a good example of the problem that is faced during the preliminary mission design phase of an asteroid mission to the teams participating to the 2nd edition of the Global Trajectory Optimisation Competition. The problem of selecting one asteroid in each of N groups containing M j possible asteroids so that a multiple rendezvous mission (appropriately named ‘asteroid grand tour’ by JPL) would spend a minimum amount of fuel / time to complete. In this paper we discuss upon this problem in general and we describe a fully automated method to select good possible asteroid sequences. Our approach starts with a branch and prune algorithm requiring a distance to be defined in the twelve dimensional space containing the orbital parameters. Once the distance is defined, the algorithm is able to select the most promising asteroid sequences in a few minutes. After that, a global optimisation problem is created for each sequence using a mixture of exponential sinusoids and Lambert’s arc representation for each trajectory leg. A combination of heuristic global optimisation solvers (Particle Swarm Optimisation, Differential Evolution, Genetic Algorithm, Simulated Annealing, Ant Colony Optimisation) is used to solve each of these global optimisation problems and to locate a good relative phasing between the asteroids.
Taking as an example the problem data assigned by JPL, we show and comment upon the results of our algorithm for different choices of the inter-orbital distance. We show that our algorithm is able to locate the correct phasing between the asteroids and returns good choices for the asteroid sequence. A final detailed optimisation is then performed to evaluate the asteroid selection quality in the real case.
- Abstract document
- Manuscript document
IAC-07-C1.7.07.pdf (🔒 authorized access only).
To get the manuscript, please contact IAF Secretariat.