Dynamics and control of a soft segmented robot for underwater and terrestrial locomotion

Abstract

The thesis investigates the modeling and control a of bio-inspired soft-segmentedrobot in underwater and terrestrial domains. Animals have evolved to navigate through complicated environments and perform tasks. Soft-bodied animals excel at manipulation in tight environments and are the inspiration for the growing research in soft robotics. With benefits including reduced weight and design complexity as compared to rigid robots, a soft segmented robot is a viable design alternative. This work explores the mathematical modeling methodology, dynamics, and control of a soft segmented robot in underwater and terrestrial domains. Planar discrete elastic rod (PDER) theory is used to model the soft-robot and simplified fluid force and friction force models are used to describe the influence of the environment. The PDER framework produces a state-space description of the dynamics that is amenable to feedback control design, as well as computationally efficient. The feedback control law is based on a decentralized control algorithm that aims to minimize the sensing and communication requirements in practice.