Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2509.20696 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918147863347200 |
|---|---|
| author | Li, Qingpeng Zhu, Chengrui Wu, Yanming Yuan, Xin Zhang, Zhen Yang, Jian Liu, Yong |
| author_facet | Li, Qingpeng Zhu, Chengrui Wu, Yanming Yuan, Xin Zhang, Zhen Yang, Jian Liu, Yong |
| contents | Enabling humanoid robots to achieve natural and dynamic locomotion across a wide range of speeds, including smooth transitions from walking to running, presents a significant challenge. Existing deep reinforcement learning methods typically require the policy to directly track a reference motion, forcing a single policy to simultaneously learn motion imitation, velocity tracking, and stability maintenance. To address this, we introduce RuN, a novel decoupled residual learning framework. RuN decomposes the control task by pairing a pre-trained Conditional Motion Generator, which provides a kinematically natural motion prior, with a reinforcement learning policy that learns a lightweight residual correction to handle dynamical interactions. Experiments in simulation and reality on the Unitree G1 humanoid robot demonstrate that RuN achieves stable, natural gaits and smooth walk-run transitions across a broad velocity range (0-2.5 m/s), outperforming state-of-the-art methods in both training efficiency and final performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_20696 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RuN: Residual Policy for Natural Humanoid Locomotion Li, Qingpeng Zhu, Chengrui Wu, Yanming Yuan, Xin Zhang, Zhen Yang, Jian Liu, Yong Robotics Enabling humanoid robots to achieve natural and dynamic locomotion across a wide range of speeds, including smooth transitions from walking to running, presents a significant challenge. Existing deep reinforcement learning methods typically require the policy to directly track a reference motion, forcing a single policy to simultaneously learn motion imitation, velocity tracking, and stability maintenance. To address this, we introduce RuN, a novel decoupled residual learning framework. RuN decomposes the control task by pairing a pre-trained Conditional Motion Generator, which provides a kinematically natural motion prior, with a reinforcement learning policy that learns a lightweight residual correction to handle dynamical interactions. Experiments in simulation and reality on the Unitree G1 humanoid robot demonstrate that RuN achieves stable, natural gaits and smooth walk-run transitions across a broad velocity range (0-2.5 m/s), outperforming state-of-the-art methods in both training efficiency and final performance. |
| title | RuN: Residual Policy for Natural Humanoid Locomotion |
| topic | Robotics |
| url | https://arxiv.org/abs/2509.20696 |