Performance and robustness comparison of energy-based methods for time-delayed systems

The removal of space debris represents one of the substantial challenges for future space missions. The development and the validation of novel technologies to tackle this problem requires a simulator capables to reproduce the space conditions on ground. For example, robotic systems can be used to simulate the dynamics of free-floating bodies in gravity-free conditions. The control of such systems is negatively affected by the presence of time delays in the control-loop and the discretization of the control inputs. These factors cause an increase in energy of the system, thence instability. In order to address the possible instability in the control of free-floating rigid bodies dynamics rendered by a robotic simulator, the thesis analyzes a simplified model of the robotic system, in which it has been calculated the minimum mass achievable on an admittance controller following a pole location analysis approach. Then three energy-based control strategies are implemented to compensate the time delay and to accurately simulate the dynamics of a free-floating body. The considered methods, namely time domain passivity approach, energy tank and wave variable, allow one to obtain system stability exploiting the passivity. The benefits and the drawbacks of each controller are evaluated in a performance and robustness study. The proposed solutions are validated in a simulation environment and experimentally on a single degree of freedom robot.