Human-Guided Multirobot Cooperative Manipulation

摘要

The interaction of a human with a team of cooperative robots, which collaboratively manipulate an object, poses significant challenges for the control design. In this paper, we propose a formation-based approach to map the human input to the motion of the object cooperatively manipulated by multiple manipulators which feature a local compliance control at the end-effector level. The formation-based approach guarantees that the reference trajectories maintain a desired geometry with respect to each other. Without being in touch with the object, the human operator is part of the formation and guides the robots explicitly. Here, the human can be interpreted as a leader in a leader-follower formation with the robotic manipulators being the followers. We analyze the system consisting of a human operator and a multirobot manipulation task in both the transient phase and the steady state for which we derive the equilibrium of the object pose from the human input and show its stability. A controllability analysis suggests that it is beneficial to make the state of the human accessible to all manipulators in order to reduce internal stress on the object. The proposed approach is evaluated in a full-scale multirobot cooperative manipulation experiment with a human.