Abstract
The cable-driven soft arm is mostly made of soft material; it is difficult to control because of the material characteristics, so the traditional robot arm modeling and control methods cannot be directly applied to the soft robot arm. In this paper, we combine the data-driven modeling method with the reinforcement learning control method to realize the position control task of robotic soft arm, the method of control strategy based on deep Q learning. In order to solve slow convergence and unstable effect in the process of simulation and migration when deep reinforcement learning is applied to the actual robot control task, a control strategy learning method is designed, which is based on the experimental data, to establish a simulation environment for control strategy training, and then applied to the real environment. Finally, it is proved by experiment that the method can effectively complete the control of the soft robot arm, which has better robustness than the traditional method.
Highlights
Compared with traditional rigid robots, soft robots are made of soft materials, which make them interact with the surrounding environment safely, so they are more suitable for the task of interaction with humans [1]
The above research focuses on the control method of the model-based soft robot arm, data-driven modeling method and reinforcement learning method of the flexible manipulator
The multi-layer perceptron (MLP) model of the soft robot arm is established; firstly, the steering gear controlled is used to traverse the motion space of the entire soft robot arm in a fixed step size, and the soft arm shape corresponding to each motion is saved by using the camera, and morphological characteristic data of the soft robot arm are extracted by image processing method
Summary
Compared with traditional rigid robots, soft robots are made of soft materials, which make them interact with the surrounding environment safely, so they are more suitable for the task of interaction with humans [1]. The above research focuses on the control method of the model-based soft robot arm, data-driven modeling method and reinforcement learning method of the flexible manipulator. In the learning process of using reinforcement learning to control the strategy, if it is directly carried out in the actual physical environment, will the driver execution process take a lot of time, but the resource consumption, such as the physical wear of the instrument, will be increased. Compared with the actual robot training, simulation training has the advantages of fast execution speed, high efficiency in learning and not requiring one to participate in the training process manually [7] Experiments show that this method can effectively achieve the expected goals, but is very meaningful to reduce resource waste and improve training efficiency.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
