Sim-to-Real Transfer for Muscle-Actuated Robots via Generalized Actuator Networks
Learns a neural actuator model from joint trajectories alone, combined with rigid-body sim, to transfer RL policies to a pneumatic-muscle arm. Claims the first sim-to-real transfer for a 4-DoF muscle-actuated arm.仅用关节位置轨迹学习神经执行器模型，结合刚体仿真，把强化学习策略迁移到气动人工肌肉机械臂上。声称首次实现 4 自由度肌肉驱动机械臂的仿真到真机迁移。
BibTeX
@misc{schneider2026sim,
title = {{Sim-to-Real Transfer for Muscle-Actuated Robots via Generalized Actuator Networks}},
author = {Jan Schneider and Mridul Mahajan and Le Chen and Simon Guist and Bernhard Schölkopf and Ingmar Posner and Dieter Büchler},
year = {2026},
eprint = {2604.09487},
archivePrefix = {arXiv},
primaryClass = {cs.RO},
url = {https://arxiv.org/abs/2604.09487},
}