Control Design for Soft Robots based on Reduced Order Model

Abstract

Inspired by nature, soft robots promise disruptive advances in robotics. Soft robots are naturally compliant and exhibit nonlinear behavior, which makes their study challenging. No unified framework exists to control these robots, especially when considering their dynamics. This letter proposes a methodology to study this type of robots around a stable equilibrium point. It can make the robot converge faster and with reduced oscillations to a desired equilibrium state. Using computational mechanics, a large-scale dynamic model of the robot is obtained and model reduction algorithms enable the design of a low-order controller and observer. A real robot is used to demonstrate the interest of the results.