Central Pattern Generator Control of a Tensegrity Based Swimmer

There has been an increased interest in the development of biologically inspired vehicles and robotics, as they have the potential of outperforming the current state of the art in metrics of efficiency, economy, maneuverability, and stealth. In marine applications, unmanned underwater vehicles currently rely on the steady state propellor or screw, which can be inefficient, noisy, application specific, and can also cause significant damage to the environment and marine life. By taking inspiration from biology, robotic systems could harness the performance of unsteady hydrodynamics for propulsion. Controlling these vehicles becomes a task of utmost importance, and the research presented in this dissertation turns again to biology for inspiration. Rhythmic motion in animals is ultimately controlled by central pattern generators (CPGs), which are systems of neurons connected in such a way that their outputs oscillate with particular phase relationships between individual neurons. It appears that these controllers produce efficient, oscillatory command signals by entraining to a resonant gait via sensory feedback. Using this approach, bio-inspired vehicles could not only capture the enhanced performance of bio-propulsion, but also robust control observed in biology. The goal of this research is to combine synthesized central pattern generators with engineered structures to demonstrate the performance capabilities of the closed loop bio-inspired system. To this end, we define a general class of tensegrity structures, derive the equations of motion, and characterize the input/output dynamics of these systems. This allows us to effectively select collocated actuators and sensors with which to integrate CPG control. The control framework is then developed and analyzed, where we investigate the mechanisms responsible for entrainment of the controller to the resonance modes of the tensegrity system. This allows us to develop