A Finite Element Method and Gaussian Process Based Digital Twin Prototype for Pneumatic Soft Actuator with Experiment Validation

The kinematic analysis of soft actuators plays a crucial role in the design and manufacturing. To address the demand for kinematic analysis of soft actuators, this paper presents a finite element methods-based simulation process for pneumatic soft actuators and validates it on a multi-chamber actuator. This study uses COMSOL Multiphysics 6.0 software and adopts the Yeoh model to describe the hyperelastic materials of pneumatic soft actuators. The finite element simulation method employed in this paper offers valuable insights for the design and improvement of pneumatic silicone soft actuators. On this basis, in order to address the issue of long computation time in finite element method, we proposed a feasible solution for the digital twin of pneumatic soft actuators by constructing a forward kinematics model of the pneumatic soft actuator using Gaussian process regression. This approach enables efficient prediction of the pneumatic soft actuator's behavior and provides a promising solution to the computational challenges in the digital twin modeling of pneumatic soft actuators.