Analysis and prospect of soft robot drive control technology

Abstract

Soft robots have become an emerging field of research with the advancement of science and technology. It has the advantages: flexible, soft and fast. It has good results within narrow space work and has great research potential in disaster relief, exploration and medical treatment. It has a theoretically infinite number of degrees of freedom, but degrees of freedom and actuators do not always correspond in a straightforward one-to-one manner, making it difficult to achieve accurate, control-accurate modelling. It is worth exploring how modelling techniques and sensor control techniques can be optimized to improve control accuracy and achieve human-machine interaction. In addition, most actuators currently use a single drive method, which leads to the limitations of redundant degree of freedom control. Combining multiple drive methods while ensuring drive control accuracy and drive force is also a future research direction. In this paper, the current state of control and drive research in soft robots will be introduced. Then three control modelling methods - modelled, modeless and hybrid control will be discussed in detail. Three common drive methods will be mentioned - fluid drive, SMA drive and EAP drive. A prospect on the control methods and drive mode of soft robots will be given. This paper is useful for robotics research.