Communication Dans Un Congrès Année : 2025

Constrained Reinforcement Learning for Unstable Point-Feet Bipedal Locomotion Applied to the Bolt Robot

Résumé

Bipedal locomotion is a key challenge in robotics, particularly for robots like Bolt, which have a point-foot design. This study explores the control of such underactuated robots using constrained reinforcement learning, addressing their inherent instability, lack of arms, and limited foot actuation. We present a methodology that leverages Constraints-as-Terminations and domain randomization techniques to enable sim-to-real transfer. Through a series of qualitative and quantitative experiments, we evaluate our approach in terms of balance maintenance, velocity control, and responses to slip and push disturbances. Additionally, we analyze autonomy through metrics like the cost of transport and ground reaction force. Our method advances robust control strategies for point-foot bipedal robots, offering insights into broader locomotion.

Fichier principal
Vignette du fichier
root.pdf (1.59 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-05198560 , version 1 (04-08-2025)

Licence

Identifiants

Citer

Constant Roux, Elliot Chane-Sane, Ludovic de Matteïs, Thomas Flayols, Jérôme Manhes, et al.. Constrained Reinforcement Learning for Unstable Point-Feet Bipedal Locomotion Applied to the Bolt Robot. 2025 IEEE-RAS 24th International Conference on Humanoid Robots (Humanoids), Sep 2025, Seoul, South Korea. pp.843-850, ⟨10.1109/Humanoids65713.2025.11203104⟩. ⟨hal-05198560⟩
966 Consultations
331 Téléchargements

Altmetric

Partager

  • More