EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Continual Reinforcement Learning for Quadruped Robot Locomotion

Sibo Gai iD, Shangke Lyu, Hongyin Zhang, Donglin Wang

Entropy · 22.01.2024 · 26 (1) · 93

DOI: 10.3390/e26010093 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Entropy, online erschienen am 22.01.2024. Verfasst von Sibo Gai und 3 weiteren Autor:innen. Beteiligte Einrichtung: School of Computer Science, Fudan University, Shanghai 200433, China. Die Arbeit wurde bisher 10-mal zitiert (Crossref) und nennt 21 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, National Science and Technology Innovation 2030, NSFC General Program.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The ability to learn continuously is crucial for a robot to achieve a high level of intelligence and autonomy. In this paper, we consider continual reinforcement learning (RL) for quadruped robots, which includes the ability to continuously learn sub-sequential tasks (plasticity) and maintain performance on previous tasks (stability). The policy obtained by the proposed method enables robots to learn multiple tasks sequentially, while overcoming both catastrophic forgetting and loss of plasticity. At the same time, it achieves the above goals with as little modification to the original RL learning process as possible. The proposed method uses the Piggyback algorithm to select protected parameters for each task, and reinitializes the unused parameters to increase plasticity. Meanwhile, we encourage the policy network exploring by encouraging the entropy of the soft network of the policy network. Our experiments show that traditional continual learning algorithms cannot perform well on robot locomotion problems, and our algorithm is more stable and less disruptive to the RL training progress. Several robot locomotion experiments validate the effectiveness of our method.

Schlagwörter

Stichwörter der Autor:innen

continual learningentropyplasticityquadruped robot locomotionreinforcement learning

Autor:innen und Institutionen

  • School of Computer Science, Fudan University, Shanghai 200433, China.
  • School of Engineer, Westlake Univercity, Hangzhou 310030, China.

ORCID-Profile: Sibo Gai

Förderung

  • National Natural Science Foundation of China
  • National Science and Technology Innovation 2030
  • NSFC General Program

Zitieren

APA

Gai, S., Lyu, S., Zhang, H., & Wang, D. (2024). Continual Reinforcement Learning for Quadruped Robot Locomotion. Entropy, 26(1), 93. https://doi.org/10.3390/e26010093

Vancouver

Gai S, Lyu S, Zhang H, Wang D. Continual Reinforcement Learning for Quadruped Robot Locomotion. Entropy. 2024;26(1):93. doi:10.3390/e26010093

BibTeX
@article{Gai2024,
  author = {Gai, Sibo and Lyu, Shangke and Zhang, Hongyin and Wang, Donglin},
  title = {Continual Reinforcement Learning for Quadruped Robot Locomotion},
  journal = {Entropy},
  year = {2024},
  volume = {26},
  number = {1},
  pages = {93},
  doi = {10.3390/e26010093}
}
Kostenlos starten →