Print Email Facebook Twitter Optimisation Strategies for Galilean Moon Tours: Low-Thrust Multiple Gravity-Assist Trajectory Design for GTOC6 Title Optimisation Strategies for Galilean Moon Tours: Low-Thrust Multiple Gravity-Assist Trajectory Design for GTOC6 Author Hoving, L. Contributor Noomen, R. (mentor) Faculty Aerospace Engineering Department Astrodynamics and Space Missions Programme Space Exploration Date 2015-12-08 Abstract Around 1610 Galileo Galilei made his discovery of the four large moons orbiting Jupiter which are referred to now as the Galilean moons. The moons Europa and Ganymede attract a significant amount of scientific interest due to potential present subsurface oceans. As consequence, the design of missions to go there and explore the moon system are increasing. This culminated in the sixth edition of the Global Trajectory Optimisation Competition (GTOC6) which is focussed on solving low-thrust multiple gravity-assist trajectories to map the Galilean moons. The aim of the thesis is to understand the complexity of the GTOC6 problem and to explore and evaluate the quality of various optimisation strategies to solve flyby sequences with low-thrust arcs. First, insight to the complexity of the problem was gained by analysing the best solution to GTOC6 so far by the Chinese Centre for Space Utilisation (CSU). From the results a clear picture was drawn from what the trajectory model should be capable of. The low-thrust trajectory model is based on the spherical shaping method that is part of the Tudat astrodynamics toolbox. A full analysis of the shaping method was performed to identify the capabilities and shortcomings of the algorithm. One of the main shortcomings is the limited accuracy for trajectories where the departure and arrival conditions differ with several degrees and more for the right ascension of the ascending node (RAAN). For optimisation use was made of differential evolution (DE). An extensive test was performed to determine the optimal settings. The result was that defining the control parameters randomly during the evolution was the best option with respect to quality and convergence. What followed was defining the optimisation model for a variable number of flybys. Furthermore, a framework was developed with six different optimisation strategies. A sequence of maximum five flybys was set to test the strategies. The strategies define the amount of freedom around the epochs of the flybys for the optimisation. Also the number of flybys that are influenced by this freedom is defined by the strategy. The goal was to optimise for ?V for a main sequence of five flybys. Here the main sequence was divided into smaller problems (subsets with less flybys). The optimisation of the main sequence was guided by the solutions of the preceding smaller subsets. Results showed that the initial subset of two flybys did not influence the optimisation of the subsequent subsets at all. Furthermore, two sequences were tested. The first sequence showed large ?V due to thrust constraint violations and limited accuracy of the spherical shaping method. On the other hand the second sequence showed ballistic solutions to go through all five moons in the sequence. Finally, from the previous test resulted an optimal strategy that was applied to a sub-problem of GTOC6. Optimisation was set to map the most interesting surfaces of the moons and to minimise ?V. The resulting trajectories were able to map the surfaces of interest. However, at the cost of more ?V compared to the previous test which only optimised for ?V. Subject low thrustshape-based methodspherical shapingmultiple gravity assistflybyinterplanetary missionGTOC6optimisation strategiesGlobal Trajectory Optimisation CompetitionDifferential EvolutionGalilean moonsmoon mappingtrajectory design To reference this document use: http://resolver.tudelft.nl/uuid:1476928c-53a3-4a3b-bb44-94d8350b7172 Part of collection Student theses Document type master thesis Rights (c) 2015 Hoving, L. Files PDF Lars_Hovings_Final_Thesis ... -11-12.pdf 19.34 MB Close viewer /islandora/object/uuid%3A1476928c-53a3-4a3b-bb44-94d8350b7172/datastream/OBJ/view