1st Global Trajectory Optimisation Competition
Solution Presented at GTOC1 by the Turin Team
In the 1st Global Trajectory Optimisation Competition multiple gravity assists could be exploited; the global optimum was to be found by choosing the correct sequence of planetary flybys in order to make the spacecraft orbit retrograde and intercept the target asteroid 2001 TW229 with a large relative velocity. Since the available time was long (30 years), a large number of planetary encounters could be performed, thus greatly reducing the propulsive effort; solutions with up to 12 flybys have been presented by other teams.
The winning solution (presented by the JPL team) has 7 gravity assist from Venus, Earth (three times), Jupiter (twice), and Saturn, uses thrust only during the first leg from Earth to Venus, and a ballistic trajectory is flown from Venus encounter to the impact with asteroid 2001 TW229. The Turin team was fifth with a 1.29 106 kg km2/s2 performance index (the winning solution performance index is 1.8 106 kg km2/s2).
The best solutions were essentially ballistic trajectories with a great number of flybys and marginal use of thrust. This kind of problems seem to be more suitable to direct methods or, better, to evolutionary algorithms, at least in order to define the flyby sequence. Such a means was not available for the Torino team (a procedure based on evolutionary algorithms for the optimisation of impulsive trajectories with multiple flyby has been developed only after the competition). The search was limited to local-optimum solutions with a reduced number (four) of flybys, that could be managed by an indirect optimisation procedure (more recently missions with up to six gravity assists have been optimised). Even though the adopted optimisation method was not apt to the proposed problem, the Turin team was able to find an accurate, (quite) simple solution with good performance. The flyby sequence (Earth, Venus, Venus, Jupiter) was suggested by simple considerations concerning astrodynamics, and the initial guess at the solution was found by means of a preliminary optimisation of simple legs that were eventually joined together. The boundary conditions at the leg extremities are chosen and progressively modified to permit the patch of the elementary legs, assuring the mission feasibility and the achievement of a high performance index.
GTOC1 Team Composition
Matteo Rosa Sentinella
Additional InformationGTOC1 problem definition
Turin team GTOC1 solution
Turin team presentation at GTOC1 workshop
© Politecnico di Torino - Corso Duca degli Abruzzi, 24 - 10129 Torino, ITALY - P.IVA 00518460019