中文标题
量化和可视化仿真到现实的差距：用于可重复性的物理引导正则化
英文标题
Quantifying and Visualizing Sim-to-Real Gaps: Physics-Guided Regularization for Reproducibility
中文摘要
使用领域随机化的仿真到现实转移方法在机器人控制中通常依赖于低齿轮比、可驱动的执行器，但当仿真到现实的差距变大时，这些方法会失效。 受传统PID控制器的启发，我们将它的增益重新解释为复杂未建模的植物动态的替代品。 然后，我们引入了一种物理引导的增益正则化方案，通过简单的现实世界实验来测量机器人有效的比例增益。 接着，在训练过程中，我们对神经控制器的局部输入输出敏感度与这些值的任何偏差进行惩罚。 为了避免朴素领域随机化的过于保守的偏差，我们还将控制器条件设置为当前的植物参数。 在一个带有110:1减速箱的现成双轮平衡机器人上，我们的增益正则化、参数条件化的RNN在硬件上的角度稳定时间与仿真结果非常接近。 同时，纯粹的领域随机化策略表现出持续的振荡和显著的仿真到现实差距。 这些结果展示了一个轻量级、可重复的框架，用于在经济的机器人硬件上缩小仿真到现实的差距。
英文摘要
Simulation-to-real transfer using domain randomization for robot control often relies on low-gear-ratio, backdrivable actuators, but these approaches break down when the sim-to-real gap widens. Inspired by the traditional PID controller, we reinterpret its gains as surrogates for complex, unmodeled plant dynamics. We then introduce a physics-guided gain regularization scheme that measures a robots effective proportional gains via simple real-world experiments. Then, we penalize any deviation of a neural controllers local input-output sensitivities from these values during training. To avoid the overly conservative bias of naive domain randomization, we also condition the controller on the current plant parameters. On an off-the-shelf two-wheeled balancing robot with a 110:1 gearbox, our gain-regularized, parameter-conditioned RNN achieves angular settling times in hardware that closely match simulation. At the same time, a purely domain-randomized policy exhibits persistent oscillations and a substantial sim-to-real gap. These results demonstrate a lightweight, reproducible framework for closing sim-to-real gaps on affordable robotic hardware.
文章页面
[2507.23445] 量化和可视化仿真到现实的差距：用于可重复性的物理引导正则化
