Closing the Reality Gap: Zero-Shot Sim-to-Real Deployment for Dexterous Force-Based Grasping and Manipulation
作者: Zhe Zhao, Zhibin Li, Yilin Ou, Mengshi Qi
分类: cs.RO
发布日期: 2026-07-06
备注: 9 pages, 9 figures. arXiv admin note: substantial text overlap with arXiv:2601.02778
💡 一句话要点
提出零-shot方法以解决多指灵巧手的现实差距问题
🎯 匹配领域: 支柱一：机器人控制 (Robot Control) 支柱二：RL算法与架构 (RL & Architecture)
关键词: 灵巧手 强化学习 触觉反馈 扭矩感知 仿真到现实转移 机器人控制 多指操作
📋 核心要点
- 现有方法在多指灵巧手的控制策略训练中面临物理交互复杂性和硬件不完美性等挑战。
- 论文提出了一种结合触觉反馈和扭矩感知的强化学习方法，旨在实现有效的模拟到现实转移。
- 实验结果显示，训练出的策略能够在真实硬件上无微调地执行可控抓取和物体重新定向，表现出良好的鲁棒性。
📝 摘要（中文）
人类般的多指灵巧手具备人类级别的操作能力，但由于接触丰富的物理特性和不完美的驱动，控制策略的训练在实际硬件上仍然困难。本文提出了一种结合密集触觉反馈和关节扭矩传感的强化学习方法，以显式调节物理交互。为实现有效的模拟到现实转移，提出了快速的触觉仿真、消除扭矩传感器需求的电流到扭矩校准，以及考虑非理想扭矩-速度效应的驱动动态建模。通过不对称的actor-critic PPO管道，完全在仿真中训练策略并直接部署到五指手上，成功实现了可控抓取力跟踪和物体重新定向等人类技能，且无需在机器人上进行微调。
🔬 方法详解
问题定义：本文旨在解决多指灵巧手在现实环境中控制策略的训练难题，现有方法在处理复杂物理交互时存在局限性，导致无法有效转移到实际硬件上。
核心思路：提出了一种结合密集触觉反馈和关节扭矩感知的强化学习框架，通过高频率、高分辨率的触觉信号来调节物理交互，从而提高策略的有效性和鲁棒性。
技术框架：整体架构包括三个主要模块：快速触觉仿真模块、扭矩校准模块和驱动动态建模模块。触觉仿真通过并行正向运动学计算虚拟触觉单元与物体之间的距离，扭矩校准则通过电流映射消除对扭矩传感器的需求。
关键创新：本文的主要创新在于首次实现了完全在仿真中训练的可控抓取策略，并成功零-shot转移到真实的多指灵巧手上，解决了传统方法在现实应用中的局限性。
关键设计：在训练过程中，采用了不对称的actor-critic PPO管道，结合触觉和扭矩信息作为观察空间，确保了策略的稳定性和高效性。
🖼️ 关键图片
📊 实验亮点
实验结果表明，训练出的策略在真实硬件上能够无微调地执行可控抓取力跟踪和物体重新定向，展现出良好的鲁棒性和适应性。这一成果标志着在多指灵巧手控制领域的重要进展。
🎯 应用场景
该研究的潜在应用领域包括机器人抓取、物体操作和人机交互等，能够为自动化生产、服务机器人及医疗辅助等行业提供更高效的解决方案。未来，随着技术的进一步发展，可能会在更复杂的操作环境中实现更广泛的应用。
📄 摘要（原文）
Human-like dexterous hands with multiple fingers offer human-level manipulation capabilities but remain difficult to train the control policies that can deploy on real hardware due to contact-rich physics and imperfect actuation. We present a sim-to-real reinforcement learning method that leverages dense tactile feedback combined with joint torque sensing to explicitly regulate physical interactions. To enable effective sim-to-real transfer, we introduce (i) a computationally fast tactile simulation that computes distances between dense virtual tactile units and the object via parallel forward kinematics, providing high-rate, high-resolution touch signals needed by RL; (ii) a current-to-torque calibration that eliminates the need for torque sensors on dexterous hands by mapping motor current to joint torque; and (iii) actuator dynamics modeling with randomization to account for non-ideal torque-speed effects and bridge the actuation gaps. Using an asymmetric actor-critic PPO pipeline, we train policies entirely in simulation and deploy them directly to a five-finger hand. The resulting policies demonstrate two essential human-hand skills: (1) command-based controllable grasp force tracking and (2) reorientation of objects in the hand, both of which are robustly executed without fine-tuning on the robot. By combining tactile and torque in the observation space with scalable sensing and actuation modeling, our system provides a practical solution to achieve reliable dexterous manipulation. To our knowledge, this is the first demonstration of controllable grasping on a multi-finger dexterous hand trained entirely in simulation and transferred zero-shot on real hardware.