arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.03067cs.RO
收藏

CMoE：对比混合专家用于人形机器人运动控制与地形适应

CMoE: Contrastive Mixture of Experts for Motion Control and Terrain Adaptation of Humanoid Robots

College of Intelligent Robotics and Advanced Manufacturing, Fudan University（智能机器人与先进制造学院，复旦大学）

Shihao Ma, Hongjin Chen, Zijun Xu, Yi Zhao, Ke Wu, Ruichen Yang, Leyao Zou, Zhongxue Gan, Wenchao Ding

阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

CMoE通过对比学习提升人形机器人在复杂地形中的运动控制与适应能力，实现专家专业化与稳健行走

摘要：

为了在现实环境中有效部署，人形机器人必须能够自主导航多样且复杂的地形，包括突变的过渡区域。尽管Vanilla混合专家（MoE）框架在理论上能够建模多样化的地形特征，但在实践中，门控网络在不同地形中表现出几乎均匀的专家激活，削弱了专家的专业化，限制了模型的表达能力。为了解决这一限制，我们引入了CMoE，一种新的单阶段强化学习框架，它整合了对比学习来优化专家激活分布。通过施加对比约束，CMoE在相同地形内最大化专家激活的一致性，同时在不同地形间最小化其相似性，从而促使专家专门针对不同的地形类型。我们通过一系列具有挑战性的实验验证了我们的方法，通过Unitree G1人形机器人进行测试。结果表明，CMoE使机器人能够连续跨越高达20厘米的台阶和高达80厘米的缝隙，同时在多样化的混合地形中实现稳健且自然的步态，超越了现有方法的极限。为了支持进一步研究并促进社区发展，我们公开发布了我们的代码。

英文摘要：

For effective deployment in real-world environments, humanoid robots must autonomously navigate a diverse range of complex terrains with abrupt transitions. While the Vanilla mixture of experts (MoE) framework is theoretically capable of modeling diverse terrain features, in practice, the gating network exhibits nearly uniform expert activations across different terrains, weakening the expert specialization and limiting the model's expressive power. To address this limitation, we introduce CMoE, a novel single-stage reinforcement learning framework that integrates contrastive learning to refine expert activation distributions. By imposing contrastive constraints, CMoE maximizes the consistency of expert activations within the same terrain while minimizing their similarity across different terrains, thereby encouraging experts to specialize in distinct terrain types. We validated our approach on the Unitree G1 humanoid robot through a series of challenging experiments. Results demonstrate that CMoE enables the robot to traverse continuous steps up to 20 cm high and gaps up to 80 cm wide, while achieving robust and natural gait across diverse mixed terrains, surpassing the limits of existing methods. To support further research and foster community development, we release our code publicly.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑