This article presents a model predictive control (MPC) method for humanoid dual-arm robots (DARs) to realize hybrid force-velocity control in the task space. The proposed method, named hybrid force-velocity MPC (HFV-MPC), enables simultaneous tracking of velocity and external force in the task space, as well as internal force regulation. First, a synchronous decoupling model of internal force, external force, and velocity is developed to address the limitations of conventional methods that decouple internal and external forces solely in the end-effector space and fail to meet the independent regulation requirements of external force and velocity in the task space. Furthermore, a generalized velocity model that integrates task-space velocity, external force, and internal force was established, and a HFV-MPC framework is constructed to enhance system robustness and achieve precise dual-arm force–velocity tracking. Subsequently, a comprehensive MPC cost function is designed, incorporating a dual-arm motion coordination coefficient to suppress undesired redundant joint motions and improve whole-body motion stability. In addition, a feedforward-linearized incremental system is derived, where the incremental model and force prediction jointly construct a generalized state-space model for MPC, to balance modeling accuracy and computational efficiency. Finally, the proposed method is validated in terms of effectiveness and robustness through both simulation and physical robotic experiments.

Visual tracking with a mobile manipulator is required to continuously keep a moving target near the center of the camera’s field of view. Conventional methods generally employ either the mobile base or the manipulator alone for target tracking. However, base-only tracking provides limited task-space motion capability, whereas arm-only tracking is restricted by the finite workspace and joint limits of the manipulator. To address this problem, this paper proposes integrating base–camera alignment and gain-scheduling visual servoing controller into a unified whole-body control frame to generate coordinated motion of the mobile base and the robotic arm for visual tracking. The mobile base and six-joint manipulator are represented as an eight-degree of freedom kinematic system, and their velocities are determined simultaneously by a quadratic program incorporating joint and velocity constraints. Auxiliary objectives regulate arm manipulability, transient motion allocation, and base–camera alignment. The proposed method enables the robot to reactively track a moving target in a three-dimensional workspace while maintaining the desired target distance. Experiments on a mobile manipulator demonstrate coordinated base–arm motion for the tested target-tracking tasks. In an illustrative controller-level step response, gain scheduling reduces the settling time from 7.99 s to 2.39 s relative to the previous fixed-gain configuration.