BriefGPT.xyz
Jul, 2023
基于扭矩的变高跳跃端到端强化学习
End-to-End Reinforcement Learning for Torque Based Variable Height Hopping
HTML
PDF
Raghav Soni, Daniel Harnack, Hauke Isermann, Sotaro Fushimi, Shivesh Kumar...
TL;DR
通过模拟到现实训练的方式，本研究提出了一种基于强化学习的力矩控制器，能够学习隐式检测相关跳跃阶段，并成功在机器人上进行实际部署。
Abstract
legged locomotion
is arguably the most suited and versatile mode to deal with natural or unstructured terrains. Intensive research into
dynamic walking
and running controllers has recently yielded great advances,
→