Implementation and analysis of adaptive control strategies for the on-orbit capture of target with a Space Manipulator System

The rapid growth of space exploration industry and the increasing need for debris mitigation missions have led to rising interest in Active Debris Removal (ADR) and On-Orbit Servicing (OOS) operations. Among robotic systems, Space Manipulator Systems (SMS) stand out for their adaptability and versatility, making them the preferred solution for OOS tasks such as satellite grasping, berthing, repair, and maintenance. Consequently, the design of proper strategies to control them have become critical. However, these control laws are often ineffective due to the strongly coupled dynamics of space systems and the difficulty in estimating the characteristics of the target. This Master's thesis aims to find new control strategies that contribute to address this problem, by proposing a robust control against the disturbances and perturbations that derive from the capture of an unknown target. A new dynamics model is proposed, in order to handle the momentum transfer during post-capture stabilization. Two different control strategies will be investigated, and their results will be compared. The first one is based on a combination of Nonlinear Dynamics Inversion (NDI) and Nonlinear Disturbance Observer (NDO), while the second one consists of NDI and Model Reference Adaptive Control (MRAC). A gain synthesis is applied to the two strategies. Overall, both the proposed methods proved capable of efficiently counteracting the disturbance introduced by the target, allowing to compensate for the perturbation and achieve a robust control of the system. However, the focus of the thesis will be on the MRAC strategy, supported by a low-pass filter to suppress high-frequency MRAC oscillations. Two different options will be provided for NDI formulation, and MRAC will be adapted accordingly. Finally, a workspace analysis will be conducted to bound some variables useful for gains synthesis, which is performed exploiting Linear Matrix Inequalities (LMI). The results achieved show how adaptive control can significantly improve the performance of space manipulator systems, paving the way for their increasingly widespread use.