A Pose Reconstruction Model of Soft Robot Arms Based on Cosserat Theory

Soft robots offer a wide range of useful applications in the fields of ground rescue, underwater grasping, space manipulation, and assistant maintenance. Pose information is important for controlling soft robotic arms; however, most pose reconstruction techniques are based on the constant curvature method. These methods are unreliable and oversimplification because they only focus on the stretching, bending, and twisting deformations of soft robotic arms. This work proposes a variable curvature pose reconstruction model based on the Cosserat theory, which involves stretching, bending, twisting, and shearing deformation to achieve accurate poses of soft robotic arms. With the help of a sensor network, it can rebuild the poses of soft robotic arms. The pose reconstruction error of the model is within 3%, according to simulation and testing data. This study offers a novel approach to soft robotic arms' pose reconstruction and improves the precision of the pose reconstruction.