arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2109.05795cs.RO
收藏

利用粗糙模拟器训练数据的软体机械臂 Q-learning 控制方法

A Q-learning Control Method for a Soft Robotic Arm Utilizing Training Data from a Rough Simulator

University of Science and Technology of China(中国科学技术大学)

Peijin Li, Gaotian Wang, Hao Jiang, Yusong Jin, Yinghao Gan, Xiaoping Chen, Jianmin Ji

AI总结:

针对软体机器人强化学习样本效率低的问题,提出借助粗糙模拟器数据预训练的 Q-learning 控制器,并在 HPN 机械臂上验证其可减少真实数据、提升精度与收敛速度。

摘要:

控制软体机器人颇具挑战,强化学习方法已在该领域得到应用并取得了可观成果。然而,由于样本效率较差,强化学习方法需要收集大量训练数据,这限制了其应用。本文提出一种用于实体软体机器人的 Q-learning 控制器,其中利用来自粗糙模拟器的数据对模型进行预训练,以提升控制器性能。我们在自研软体机器人,即 Honeycomb Pneumatic Network(HPN,蜂窝气动网络)机械臂上实现了该方法。实验表明,使用预训练模型不仅可以减少真实世界训练数据的数量,还能大幅提升其精度和收敛速度。

英文摘要:

It is challenging to control a soft robot, where reinforcement learning methods have been applied with promising results. However, due to the poor sample efficiency, reinforcement learning methods require a large collection of training data, which limits their applications. In this paper, we propose a Q-learning controller for a physical soft robot, in which pre-trained models using data from a rough simulator are applied to improve the performance of the controller. We implement the method on our soft robot, i.e., Honeycomb Pneumatic Network (HPN) arm. The experiments show that the usage of pre-trained models can not only reduce the amount of the real-world training data, but also greatly improve its accuracy and convergence rate.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑