Dynamic Legged Ball Manipulation on Rugged Terrains with Hierarchical Reinforcement Learning
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866909586294833152 |
|---|---|
| author | Zhu, Dongjie Yang, Zhuo Wu, Tianhang Ge, Luzhou Li, Xuesong Liu, Qi Li, Xiang |
| author_facet | Zhu, Dongjie Yang, Zhuo Wu, Tianhang Ge, Luzhou Li, Xuesong Liu, Qi Li, Xiang |
| contents | Advancing the dynamic loco-manipulation capabilities of quadruped robots in complex terrains is crucial for performing diverse tasks. Specifically, dynamic ball manipulation in rugged environments presents two key challenges. The first is coordinating distinct motion modalities to integrate terrain traversal and ball control seamlessly. The second is overcoming sparse rewards in end-to-end deep reinforcement learning, which impedes efficient policy convergence. To address these challenges, we propose a hierarchical reinforcement learning framework. A high-level policy, informed by proprioceptive data and ball position, adaptively switches between pre-trained low-level skills such as ball dribbling and rough terrain navigation. We further propose Dynamic Skill-Focused Policy Optimization to suppress gradients from inactive skills and enhance critical skill learning. Both simulation and real-world experiments validate that our methods outperform baseline approaches in dynamic ball manipulation across rugged terrains, highlighting its effectiveness in challenging environments. Videos are on our website: dribble-hrl.github.io. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_14989 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Dynamic Legged Ball Manipulation on Rugged Terrains with Hierarchical Reinforcement Learning Zhu, Dongjie Yang, Zhuo Wu, Tianhang Ge, Luzhou Li, Xuesong Liu, Qi Li, Xiang Robotics Advancing the dynamic loco-manipulation capabilities of quadruped robots in complex terrains is crucial for performing diverse tasks. Specifically, dynamic ball manipulation in rugged environments presents two key challenges. The first is coordinating distinct motion modalities to integrate terrain traversal and ball control seamlessly. The second is overcoming sparse rewards in end-to-end deep reinforcement learning, which impedes efficient policy convergence. To address these challenges, we propose a hierarchical reinforcement learning framework. A high-level policy, informed by proprioceptive data and ball position, adaptively switches between pre-trained low-level skills such as ball dribbling and rough terrain navigation. We further propose Dynamic Skill-Focused Policy Optimization to suppress gradients from inactive skills and enhance critical skill learning. Both simulation and real-world experiments validate that our methods outperform baseline approaches in dynamic ball manipulation across rugged terrains, highlighting its effectiveness in challenging environments. Videos are on our website: dribble-hrl.github.io. |
| title | Dynamic Legged Ball Manipulation on Rugged Terrains with Hierarchical Reinforcement Learning |
| topic | Robotics |
| url | https://arxiv.org/abs/2504.14989 |