Position synchronization through the energy-reflection based time domain passivity approach in position-position architectures

Abstract

Position synchronization is a primary control challenge in the delayed coupling of robotic systems. While in tele-operation, the interaction force of the robot is often the preferred feedback for high degrees of transparency, positions are exchanged in the coupling of autonomous systems for improved position matching. This letter proposes the extension of the energy-reflection based Time Domain Passivity Approach to position-position control for delayed coupled network systems. By considering the physical behavior of the two coupling springs as a reference for the desired energy exchange of the agents, a real passive coupling 2-port is achieved at low conservatism. Furthermore, position drift is completely avoided in contrast to state-of-the-art approaches. The high modularity of the approach allows a straightforward application in multilateral setups. Other benefits are its robustness in case of active environments and an undisturbed force transmission to the input device in tele-operation setups. The approach is validated in bilateral and multilateral experiments with delays up to 800 ms and compared to a state-of-the-art approach. The results confirm the high position tracking and force-feedback quality of the proposed approach as well as its robustness against active environments.