EUVIMEDEUVIMED entdecken
Crossref

Reinforcement learning for versatile, dynamic, and robust bipedal locomotion control

Zhongyu Li, Xue Bin Peng, Pieter Abbeel, Sergey Levine, Glen Berseth, Koushil Sreenath

The International Journal of Robotics Research · 2025-04 · 44 (5) · 840-888

DOI: 10.1177/02783649241285161 ↗

Kurz gefasst

Fachartikel in der Zeitschrift The International Journal of Robotics Research, online erschienen am 23.10.2024. Verfasst von Zhongyu Li und 5 weiteren Autor:innen. Die Arbeit wurde bisher 142-mal zitiert (Crossref) und nennt 111 Literaturangaben. Gefördert von: The AI Institute, National Science Foundation, Canadian Institute for Advanced Research.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper presents a comprehensive study on using deep reinforcement learning (RL) to create dynamic locomotion controllers for bipedal robots. Going beyond focusing on a single locomotion skill, we develop a general control solution that can be used for a range of dynamic bipedal skills, from periodic walking and running to aperiodic jumping and standing. Our RL-based controller incorporates a novel dual-history architecture, utilizing both a long-term and short-term input/output (I/O) history of the robot. This control architecture, when trained through the proposed end-to-end RL approach, consistently outperforms other methods across a diverse range of skills in both simulation and the real world. The study also delves into the adaptivity and robustness introduced by the proposed RL system in developing locomotion controllers. We demonstrate that the proposed architecture can adapt to both time-invariant dynamics shifts and time-variant changes, such as contact events, by effectively using the robot’s I/O history. Additionally, we identify task randomization as another key source of robustness, fostering better task generalization and compliance to disturbances. The resulting control policies can be successfully deployed on Cassie, a torque-controlled human-sized bipedal robot. This work pushes the limits of agility for bipedal robots through extensive real-world experiments. We demonstrate a diverse range of locomotion skills, including: robust standing, versatile walking, fast running with a demonstration of a 400-meter dash, and a diverse set of jumping skills, such as standing long jumps and high jumps.

Förderung

  • The AI Institute
  • National Science Foundation
  • Canadian Institute for Advanced Research

Zitieren

APA

Li, Z., Peng, X. B., Abbeel, P., Levine, S., Berseth, G., & Sreenath, K. (2025). Reinforcement learning for versatile, dynamic, and robust bipedal locomotion control. The International Journal of Robotics Research, 44(5), 840-888. https://doi.org/10.1177/02783649241285161

Vancouver

Li Z, Peng XB, Abbeel P, Levine S, Berseth G, Sreenath K. Reinforcement learning for versatile, dynamic, and robust bipedal locomotion control. The International Journal of Robotics Research. 2025;44(5):840-888. doi:10.1177/02783649241285161

BibTeX
@article{Li2025,
  author = {Li, Zhongyu and Peng, Xue Bin and Abbeel, Pieter and Levine, Sergey and Berseth, Glen and Sreenath, Koushil},
  title = {Reinforcement learning for versatile, dynamic, and robust bipedal locomotion control},
  journal = {The International Journal of Robotics Research},
  year = {2025},
  volume = {44},
  number = {5},
  pages = {840--888},
  doi = {10.1177/02783649241285161}
}
Kostenlos starten →