Not Only Rewards but Also Constraints: Applications on Legged Robot Locomotion

Several earlier studies have shown impressive control performance in complex robotic systems by designing the controller using a neural network and training it with model-free reinforcement learning. However, these outstanding controllers with natural motion style and high task performance are devel...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	IEEE transactions on robotics 2024, Vol.40, p.2984-3003
Hauptverfasser:	Kim, Yunho, Oh, Hyunsik, Lee, Jeonghyun, Choi, Jinhyeok, Ji, Gwanghyeon, Jung, Moonkyu, Youm, Donghoon, Hwangbo, Jemin
Format:	Artikel
Sprache:	eng
Schlagworte:	Algorithms Constrained reinforcement learning (RL) Controllers Legged locomotion Locomotion Machine learning Neural networks Optimization Quadrupedal robots Reinforcement learning Robot dynamics Robot learning Robots Training
Online-Zugang:	Volltext bestellen
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!