A Q-learning Control Method for a Soft Robotic Arm Utilizing Training Data from a Rough Simulator
This work addresses the problem of sample inefficiency in reinforcement learning for soft robotics, offering a domain-specific incremental improvement.
The paper tackles the challenge of controlling a soft robot by proposing a Q-learning controller that uses pre-trained models from a rough simulator to reduce real-world training data needs, resulting in improved accuracy and convergence rates.
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.