Model-based Control for a Coupled Soft Robotic Exo-digit and Anthropomorphic Finger with Multi-Contact Point Physical Interaction

Abstract

A soft robotic exoskeleton digit can be a prosthetic implant, a solution for finger/toe loss, or an element of rehabil-itation gloves for the human hand's functional recovery. This paper presents the modeling and feedback linearization control of the quasi-static motion of a soft robotic exo-digit physically interacting with an anthropomorphic finger model for tracking the desired trajectory. The exo-digit is modeled as an articulated hybrid soft-and-rigid model with constant bending curvatures and arc lengths. Considering the soft elastomer property of the silicone, a hyperelastic constitute model based on Yeoh's 3<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">rd</sup>-order material model is used. The anthropomorphic finger is modeled as an rigid multi-body structure with integrated torsional springs at its joints. A coupled kinematics and quasi-static motion is developed to model the multi-contact points interaction between the robot and finger model. This model is then presented as a nonlinear discrete-time MIMO state-space system for the control system design of the physical human-robot interaction. The input-output feedback linearization is utilized, and a control law is designed. Here, the control input is the actuation pressure of each independent actuation point, and the output is the pose of the anthropomorphic model's tip. The theoretical model and trajectory tracking control law were experimentally and numerically examined; the results show the effectiveness of the model and control law in tracking the desired fingertip pose trajectory.