ESA GNC Conference Papers Repository

In the framework of Earth Observation missions, many space programmes have been focused on remote sensing the Earth, to observe events and phenomena which cannot be studied in-situ. A key role in this context is played by autonomous guidance systems. From this point of view, Nonlinear Model Predictive Control owns a great potential for the future of aerospace control and guidance systems thanks to its capability to accomplish jointly the guidance and control tasks, ensuring the minimization of a suitable performance index. In this paper, a novel Nonlinear Model Predictive Control framework for autonomous guidance and control with high-thrust quasi-impulsive maneuvers is presented. A key feature is the use of different kinds of orbital motion models as possible internal prediction models in the optimization algorithm. These models are based on: Cartesian Coordinates, Keplerian Orbital Elements and Modified Equinoctial Orbital Elements. The ESA Sentinel-2 mission is considered as a benchmark for the proposed framework. The obtained results show the effectiveness, in terms of propellant consumption and reference tracking, of using the Equinoctial Elements as internal model. This latter dynamics parametrization can also overcome possible singularities (affecting the Keplerian Elements) while reducing the computational complexity.