Review of design and modeling of dielectric elastomer actuators

Natural muscles provide animals with good athletic ability. As the basic unit of movement, skeletal muscle fibers can make the skeleton produce linear or rotational movement. Inspired by natural muscles, researchers have developed various flexible actuators (artificial muscles) based on dielectric elastomer materials to drive flexible robots. Compared with the conventional rigid robots driven by motors, the flexible robots driven by dielectric elastomer actuators have unique advantages, especially in terms of interaction with humans. An integrated drive-structure design makes the robot realize save human-robot integration. According to the types of output motions, this paper first summarizes the existing design, modeling methods and applications of various dielectric elastomer actuators. As the electromechanical model obtained by the energy methods is not accurate, a data-driven approach based on the Gaussian Process Regression (GPR) is proposed to compensate the positioning errors resulting from the inaccurate model. The work presented in this paper has certain significance for the design and modeling of dielectric elastomer actuators.