发表机构
University of Michigan; University of California, Los Angeles; Amazon; California Institute of Technology（密歇根大学; 加利福尼亚大学洛杉矶分校; 亚马逊; 加州理工学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对人形机器人搬运重物时的全身强力交互难题，提出HULK框架，结合MPC引导强化学习与捕获点控制障碍函数训练双教师策略并蒸馏为单一策略，显著降低跟踪误差和DCM偏移，并在仿真和Unitree G1上验证了有效性。
AI 中文摘要
人形机器人对大型、重型物体的移动操作需要全身的强力交互。然而，此类负载会改变人形机器人的质心，并在上半身施加持续载荷，对平衡和指令跟踪构成挑战。我们提出了HULK，一个用于强力移动操作的全身控制框架。利用模型预测控制（MPC）来引导强化学习，并预测负载动力学，我们训练了两个教师策略：一个在腕部受力下跟踪手臂运动，另一个在将大型物体紧贴身体时进行移动。一个捕获点控制障碍函数在训练期间增强腕部受力教师策略，以改善负载下的平衡。我们将两个教师策略蒸馏成一个统一策略。评估涵盖仿真和Unitree G1。在仿真中，带有障碍函数的教师策略在每臂10 kg负载下，在前向和侧向速度跟踪误差上均达到评估控制器中的最低值，并将聚合发散分量（DCM）偏移幅度相对仅用MPC引导的强化学习降低了35.7%。我们的腕部受力教师策略能够承受高达130 N的躯干推力扰动。
英文摘要
Humanoid loco-manipulation of large, heavy objects demands forceful interaction across the entire body. However, such payloads shift a humanoid's center of mass and impose sustained loads across the upper body, challenging balance and command tracking. We present HULK, a whole-body control framework for forceful loco-manipulation. Using model predictive control (MPC) to guide reinforcement learning with predictions of the loaded dynamics, we train two teachers: one tracks arm motions under wrist forces, and the other locomotes while holding large objects against the body. A capture-point control barrier function augments the wrist-force teacher during training to improve balance under load. We distill both teachers into a single policy. Evaluation spans simulation and the Unitree G1. In simulation, the teacher with the barrier function achieves the lowest forward and lateral velocity tracking errors at 10 kg per arm among evaluated controllers and reduces aggregate divergent component of motion (DCM) excursion magnitude by 35.7% relative to MPC-guided reinforcement learning alone. Our wrist-force teacher withstands torso push disturbances of up to 130 N.
Comments16 pages, 7 figures, IEEE International Conference on Robotics and Automation 2027
