Design and Testing of Guidance Strategies for In-Orbit Servicing and Small Proximity Operations

In the last decade, In-orbit servicing and active debris removal gained a lot of interests in the scientific community, both for the need of keeping the LEO orbit accessible and for the economic reasons that the implementation of this kind of missions can bring. In fact, the aims of in-orbit servicing are to repair, refuel, upgrade, de-orbit or extend the life of satellites, making missions last longer. Therefore, debris presence around Earth orbit will be reduced, diminishing the risks for space missions. In this scenario, to safely accomplish mission’s goals, it’s crucial to implement innovative GNC techniques that can be considered reliable and adaptable to different scenarios. It’s key to develop techniques for collaborative known satellites and even for non collaborative ones, in order to perform in-orbit servicing for new and older satellites. Furthermore, another aspect to be considered is the growing and promising implementation of artificial intelligence, which can enhance and enable innovative controller approaches. As a result, the aim of this thesis is to implement an orbit simulator for rendezvous missions in circular LEO orbit, able to adapt to different satellite architectures, and to investigate two different path planning methods to be used in small proximity operations. In addition, to make the simulation adaptable to different missions, it was decided to implement an initial sensing phase, outside the approach ellipsoid, in which the chaser searches for the target and then reaches the position to begin proximity manoeuvres The first type of guidance investigated consists of an innovative relative circular IFDS trajectory used for mapping an unknown object or to estimate its position and attitude. The second one consists of an APF guidance suitable for guiding the chaser from any position inside the approach ellipsoid into a holding point outside the Keep-Out zone, hence maintaining a safety distance of 200 [m] from the target. After reaching the holding point, the chaser will continue to approach the target through a final path into the safety cone. The project is implemented in Matlab/Simulink, Blender is also used to gather images of the target during the proximity operations. The simulations show that the 2 guidance strategies allow maintaining safety standards independently to initial conditions, even though time or propeller consumption optimization was not considered. This aspect can represent the next phase of the project, which will focus on using AI to control the spacecraft.