Going in Circles Revolutionizing Repetitive Control, with Application to Tethered Energy Systems
Seminar | September 19 | 10-11 a.m. | 3110 Etcheverry Hall
Associate Professor Chris Vermillion, North Carolina State University
Abstract: Repetitive processes, including manufacturing operations involving thousands of copies of a part, control of hard disk drives, and tracking of periodic gaits in human exoskeletons, have represented a significant area of automatic control over the past several decades. While most repetitive control theory focuses on classical design goals of periodic setpoint tracking and disturbance rejection, there exist a number of recent applications for which the ultimate goal of repetitive control is not merely to track a prescribed reference signal (or reject a periodic disturbance) but instead to maximize or minimize some economic metric. For example, the controller for a pick-and-place robot ultimately should be designed to minimize task execution time. An actively controlled exoskeleton should be designed to minimize a cost function related to metabolic cost. A wind energy system should be designed to maximize average net power output. This talk will focus on the development of economic repetitive control strategies, which will aim to use iteration-to-iteration learning to minimize (or maximize) a profitability index, rather than merely focusing on tracking prescribed setpoints and rejecting prescribed disturbances. The talk will focus specifically on tethered energy systems, which are wind and marine hydrokinetic energy systems where the conventional tower is replaced with tethers and a lifting body (a rigid wing or kite). It can be shown that flying the lifting body in repetitive (circular or elliptical) patterns can dramatically increase power production over stationary flight. This talk will show how economic repetitive control tools can be used to optimize the power output of these revolutionary energy systems in the presence of varying environmental conditions.
Biography: Chris Vermillion received his Ph.D. in Electrical Engineering from the University of Michigan in 2009 and received his undergraduate degrees in Aerospace and Mechanical Engineering from the University of Michigan in 2004. Immediately following his Ph.D. work, Dr. Vermillion worked on advanced automotive powertrain control, focusing on constrained optimal control approaches that simultaneously addressed the competing performance interests of fuel economy, emissions, drivability, and torque delivery. Subsequently he served as a Lead Engineer for Altaeros Energies and managed all of the dynamic modeling, control system design, software development, and embedded hardware development for Altaeros lighter-than-air wind energy system. Dr. Vermillion has participated in the full-scale fight testing of two of Altaeros designs. Dr. Vermillion is currently an Associate Professor at NC State, where his research focuses on the dynamic characterization, design optimization, and optimal control of airborne wind energy systems, marine hydrokinetic energy systems, and energy-efficient connected and autonomous vehicles. Dr. Vermillion was the recipient of the National Science Foundations CAREER Award in 2015, the UNC-Charlotte Maxheim Research Fellowship in 2016, and the UNC-Charlotte College of Engineering Excellence in Teaching Award in 2017.