外观
论文arXiv 2608.102202 个月前
基于自碰撞避免参考的人形机器人狭窄空间导航全身规划
针对人形机器人狭窄空间导航，提出结合可达刚体体积与可微碰撞避免的三阶段全身规划，并在Unitree G1基准与物理仿真中验证。
宇树 G1原文链接：Whole-Body Planning for Humanoids Navigating Confined Spaces via Self-Collision Avoidance References（arXiv 2608.10220 · 发布日期：2026-08-10）
Whole-Body Planning for Humanoids Navigating Confined Spaces via Self-Collision Avoidance References
针对人形机器人狭窄空间导航，提出结合可达刚体体积与可微碰撞避免的三阶段全身规划，并在Unitree G1基准与物理仿真中验证。
核心要点
- 高度受限环境需同时处理密集障碍、自碰撞边界和多接触动态可行性。
- 框架直接在运动学可达刚体体积上规划路径，生成体积感知引导。
- 三个基准超过NIST应急响应标准，方法在基线失败处生成12至18秒可行轨迹。
- 优化计划用于训练残差强化学习策略，在物理仿真中经域随机化成功跟踪。
验证范围与局限：验证包括Unitree G1的三个基准测试床和物理仿真中的策略跟踪。
英文摘要 (Abstract)
Humanoid locomotion in highly confined environments requires navigating dense environmental obstacles and complex self-collision bounds while maintaining multi-contact dynamic feasibility. Traditional trajectory optimizers frequently struggle in these restricted spaces, as navigating the large collision space with splines on particle abstractions is insufficient and leads to poor local minima. To address this, we propose a three-stage whole-body planning framework that formulates kinematic path planning directly over kinematically reachable rigid-body volumes. By integrating differentiable collision avoidance into a reachability-constrained formulation, our framework synthesizes volume-informed guides that reliably guide a full-order trajectory optimizer over long horizons. We show that these optimized plans serve as high-quality references to train a residual reinforcement learning policy for robust online execution. We validate our approach on the Unitree G1 humanoid across three benchmark testbeds exceeding NIST emergency response standards, achieving restricted confinement ratios ($C_r < 1.5$). Our framework generates feasible trajectories across 12-to-18-second tasks with complex foot and hand contacts where standard baselines fail, while the learned policy successfully tracks these plans under extensive domain randomization in physics simulation.
作者：Carlos González, Luis Sentis
来源
本页事实对应的公开页面，以原文为准。