Human-Inspired Holistic Control for Mobile Humanoid Robots

Humanoid mobile manipulators integrate a humanoid upper body with a mobile platform, forming a highly redundant system capable of performing complex manipulation tasks. To address the redundancy arising from the coordinated motion of the wheeled base, waist, and dual arms, this study proposes a human-inspired holistic control method based on multi-objective optimization. The degrees of freedom (DOF) of the upper limbs and the mobile base are unified within a single control framework, thereby enhancing overall motion coordination. Specifically, the controller is formulated as a strictly convex quadratic program (QP) that ensures accurate end-effector tracking while effectively handling joint position and velocity constraints. Inspired by human motor characteristics, the method incorporates a hierarchical weight assignment strategy and base DOF optimization to preserve arm manipulability while achieving effective coordination between the base and waist. Simulation studies of dual-arm handling tasks and real-world experiments involving mobile handling and peg-in-hole assembly demonstrate that the proposed method generates smooth, humanoid-like motions, thereby validating the effectiveness of the proposed control framework.