Human-Robot Cooperative Manipulation Based on MPC and Whole-body Control

Abstract

The mobile manipulator combines the high precision of the manipulator with the mobility of the mobile base, finding widespread applications in intelligent logistics, manufacturing, and home services. Mobile manipulators face challenges in coordinating control between the manipulator and the base and an immature human-robot collaboration mechanism. To address these issues, this paper proposes a human-robot collaborative manipulation framework comprising a collaborative controller and a whole-body controller. The collaborative controller is based on admittance control and incorporates dynamically adjustable damping. It enables the robot to exert low traction forces during object transportation and increases damping as the robot approaches the target position, ensuring smooth stopping and improved stability during collaboration. The whole-body controller ensures the coordinated motion performance of the mobile manipulator. The reliability and stability of the framework are validated through three human-robot collaborative manipulation experiments.