06 Fakultät Luft- und Raumfahrttechnik und Geodäsie
Permanent URI for this collectionhttps://elib.uni-stuttgart.de/handle/11682/7
Browse
1 results
Search Results
Item Open Access Optimization of the orbit parameters of future gravity missions using genetic algorithms(2011) Ellmer, MatthiasThe best global models for Earth's gravitational field are provided by recent satellite missions like CHAMP, GRACE and GOCE. The research in this thesis is dedicated to finding favorable parameters for possible future multi-pair satellite formations that follow in GRACE's footsteps. A software framework was developed in the Python programming language that allows the study of such satellite formations using a range-acceleration approach to gravity recovery. The computational performance of this framework was improved significantly through several optimization steps, amongst others involving successful parallelization of multiple simulations and the exploitation of a GPU-accelerated version of the mathematical library BLAS. A large number of parameter studies were performed in an attempt to define criteria for the automatic quantification of the objective quality of specific multi-pair mission scenarios. This was achieved by analyzing and assessing several characteristics of the recovered potentials in both the spatial and the spectral domain. Using the custom software framework as well as the developed benchmarks for a recovered potential, a genetic algorithm capable of autonomously finding optimal combinations of multiple satellite formations was implemented. This genetic algorithm was then tested in a simple scenario, intended to find the optimal complimentary formation to a single polar satellite pair modeled after the GRACE twin-satellite mission. The performed genetic algorithm simulation suggests that a pendulum on a ~58° inclined orbit with a long along-track baseline and an opening angle of around 40° would make a good companion for GRACE. While finalizing this thesis a flaw in the custom software was discovered that affected all simulations performed in the context of the parameter studies and the genetic algorithm. This means that the results presented in this thesis should be considered with a critical mindset. The developed methodologies used to arrive at these results are however sound and could be used in future studies.