web analytics
Site logo

Dr. Tansel Yucelen


Laboratory for Autonomy, Control, Information, and Systems


Verifiable Adaptive Control Systems and Learning Algorithms


2019 IEEE Conference on Decision and Control Full-Day Workshop
December 10, 2019 - Nice, France




A. Organizer and Co-Organizers

Tansel Yucelen, University of South Florida, yucelen@usf.edu (Organizer)
Anuradha Annaswamy, Massachusetts Institute of Technology, aanna@mit.edu (Co-Organizer)
Warren Dixon, University of Florida, wdixon@ufl.edu (Co-Organizer)
K. Merve Dogan, University of South Florida, dogan@mail.usf.edu (Co-Organizer)
Jonathan A. Muse, Air Force Research Laboratory, jonathan.muse.2@us.af.mil (Co-Organizer)
Frank Lewis, University of Texas Arlington, lewis@uta.edu (Co-Organizer)


B. Objective and Significance

Behavior of physical systems are generally perturbed due to the presence of exogenous disturbances (e.g., resulting from winds and turbulence) and system uncertainties (e.g., resulting from imperfect modeling, degraded modes of operation, changes in dynamics, damaged control surfaces, and sensor failures). As a consequence, a fundamental problem in the design of their feedback control architectures is to achieve closed-loop system stability, performance, and robustness against exogenous disturbances and system uncertainties. Unlike fixed-gain control architectures, adaptive control systems offer the capability to deal with exogenous disturbances and system uncertainties, in an online fashion, through learning. This implies that they are not tuned to a worst-case scenario and they continuously improve their performance in real-time. These two appealing aspects make adaptive control systems and learning algorithms important candidates for a wide array of physical systems. Although government and industry agree on their potential in providing vehicle safety and reducing vehicle development costs, a major issue is the lack of system-theoretic methods for their verification, due to their nonlinear nature. Motivated by this standpoint, the objective of this full-day workshop is to cover the state-of-the-art verifiable system-theoretic approaches in adaptive control systems and learning algorithms for their safe and reliable real-world applications.

This workshop will start with an introductory talk on the preliminaries in adaptive control systems and learning algorithms (see below
Talk 1 in Section C), where it will be followed by 6 related talks. Specifically, we will provide novel system-theoretic methods for achieving transient and steady-state performance guarantees with adaptive control systems (see below Talk 2 in Section C). These methods will have the capability to assign a-priori given, user-defined performance bounds on the adaptively controlled closed-loop system trajectories to allow for their verifiable implementations in real-world applications. In adaptive control of physical systems, it is well-known that actuator (e.g., slow versus fast actuation) and unmodeled dynamics (e.g., the presence of flexible dynamics and appendages connected to rigid bodies) in feedback loops can yield to unstable closed-loop system trajectories. To address this fundamental gap, our workshop includes talks on how a control engineer can design and analyze adaptive control systems in the presence of actuator and unmodeled dynamics (see below Talks 4 and 5 in Section C respectively), where fundamental stability verification conditions will be covered toward robust real-time execution of adaptive control systems and learning algorithms. We will also focus on the state-of-the-art approaches to enable the use of adaptive control algorithms for physical systems with switching modes (see below Talk 6 in Section C). Advances in analysis methods that allow arbitrary switching between asymptotic stable subsystems will be described along with new adaptive update laws to facilitate verifiable dwell time conditions for switching between stable and unstable subsystems.

Accelerated learning techniques have seen widespread use in machine learning and have led to parameter convergence through the use of stochastic inputs. Parameter estimation methods have long been examined in adaptive control systems as a central learning component, with emphasis on necessary and sufficient conditions for convergence and fast convergence rates. Within the scope of our workshop, we will examine concepts in stability, asymptotic convergence, and learning, common to both machine learning and adaptive control systems (see below
Talk 3 in Section C). Building on the similarities in update laws and common concepts, new intersections and opportunities for improved algorithm analysis for fast parameter estimation and accelerated learning in physical systems are provided. In particular, integral-based accelerated learning algorithms for adaptive control will be presented as they relate to accelerated parameter convergence. Finally, we will cover recent discoveries in neurocognitive psychology about how the brain functions provide new feedback control structures known as integral reinforcement learning (IRL) for learning online the solutions to optimal control problems and multi-player differential games (see below Talk 7 in Section C). IRL yields adaptive feedback control systems with better properties, greater intelligence, and optimal performance. We will present IRL applications in human-robot interaction, industrial process control, unmanned aerial vehicle control, and elsewhere. New classes of reinforcement learning adaptive control algorithms will be developed and rigorous proofs of convergence will verify their stability, performance, and robustness.

The
significance of this workshop will be the coverage of the state-of-the-art methods in adaptive control systems and learning algorithms for their verifiable real-world implementation. We will also cultivate new future research directions under a 90 minutes long panel discussion involving organizers and expected workshop attendees.


C. Schedule

09:00 - 09:45: Tansel Yucelen (Talk 1)
Preliminaries in Adaptive Control Systems and Learning Algorithms

09:45 - 10:30: Tansel Yucelen (
Talk 2)
Adaptive Control Architectures with Verifiable Performance Guarantees

10:30 - 10:45: Coffee Break

10:45 - 11:30: Anuradha Annaswamy (
Talk 3)
Algorithms for Accelerated Learning in Machine Learning and Fast Parameter Convergence in Adaptive Control Systems

11:30 - 12:15: Jonathan A. Muse (
Talk 4)
Verifiable Adaptive Control Architectures for Uncertain Systems with Actuator Dynamics

12:15 - 13:30: Lunch

13:30 - 14:15: K. Merve Dogan (
Talk 5)
Relaxing Stability Limit of Adaptive Control Architectures Against Unmodeled Dynamics

14:15 - 15:00: Warren Dixon (
Talk 6)
Control-Theoretic Foundations of Switched Adaptive Systems

15:00 - 15:15: Coffee Break

15:15 - 16:00: Frank Lewis (
Talk 7)
Verifiable Reinforcement Learning in Control Applications

16:00 - 17:30: Panel Discussion


D. Audience

Although government and industry agree on the potential of adaptive control systems and learning algorithms, a major issue is the lack of system-theoretic methods for their verification. Motivated by this standpoint, this workshop will enable individuals from academia, government, and industry to learn about the state-of-the-art approaches in adaptive control systems and learning algorithms for their safe and reliable real-world implementations. Specifically, the presenters of this workshop will cover topics addressing how to implement adaptive control systems with verifiable transient and steady-state performance guarantees, how to address the presence of actuator and unmodeled dynamics when adaptive control systems are in feedback loops, how to design and analyze adaptive control systems for physical plants with switching modes, and how to advance adaptive control systems with system-theoretic guarantees using tools and methods from machine and reinforcement learning. This workshop will be relevant to practicing professionals from electrical, mechanical, and aerospace industries. It also intends to cultivate new future research directions under a panel discussion involving organizers and expected workshop attendees. Finally, this workshop is expected to be a great value to experts and students in the adaptive control systems and learning algorithms fields.



Tansel Yucelen is an Assistant Professor in the Department of Mechanical Engineering at the University of South Florida (since 2016). He received the Doctor of Philosophy degree in Aerospace Engineering from the Georgia Institute of Technology (2012). Prior to joining the University of South Florida, he held an Assistant Professor position in the Department of Mechanical and Aerospace Engineering at the Missouri University of Science and Technology (2013-2016) and a Research Engineer position in the School of Electrical and Computer Engineering and the School of Aerospace Engineering at the Georgia Institute of Technology (2011-2013). He was also a Summer Faculty Fellow at the Air Force Research Laboratory Wright-Patterson (2014) and Eglin (2015), and a consultant to NASA (2014-2016), Wichita State University (2017-2018), and the Missouri University of Science and Technology (2017-2018). He has co-authored more than 200 peer-reviewed papers in top internationally-recognized journals and conferences, and secured external grants and contracts through NSF, AFRL, AFOSR, ARO, DARPA, MDA, NASA, and ORAU. He was the recipient of the University of South Florida Research and Innovation Faculty Outstanding Research Achievement Award (2018), the University of South Florida College of Engineering Junior Outstanding Research Achievement Award (2017), the Aerospace Control and guidance Systems Committee Dave Ward Memorial Lecture Award (2016), the AIAA Technical Contribution Award (2016), the Oak Ridge Associated Universities Junior Faculty Award (2015), and the Class of 1942 Excellence in Teaching Award (2014). He is a member of the National Academy of Inventors, a senior member of the AIAA, and a senior member of the IEEE.

Anuradha Annaswamy is Founder and Director of the Active-Adaptive Control Laboratory in the Department of Mechanical Engineering at MIT. She is recognized worldwide as a pioneer in adaptive control theory and its applications to aerospace, automotive, and propulsion systems as well as cyber physical systems such as Smart Grids, Smart Cities, and Smart Infrastructures. Her current research team of 15 students and post-docs is supported by Air-Force Research Laboratory, Boeing, Ford-MIT Alliance, Department of Energy, and NSF. Dr. Annaswamy is an author of over 100 journal publications and 250 conference publications, co-author of a graduate textbook on adaptive control, and co-editor of several cutting edge science and technology reports including Systems & Control for the future of humanity, research agenda: Current and future roles, impact and grand challenges (Annual Reviews in Control, 2016), Smart Grid Control: Overview and Research Opportunities (Springer, 2018), and Impact of Control Technology (IoCT-report 2011 and 2013). Dr. Annaswamy has received several awards including the George Axelby (1986) and Control Systems Magazine (2010) best paper awards from the IEEE Control Systems Society (CSS), the Presidential Young Investigator award from NSF (1992), the Hans Fisher Senior Fellowship from the Institute for Advanced Study at the Technische Universita╠łt Mu╠łnchen (2008), the Donald Groen Julius Prize from the Institute of Mechanical Engineers (2008). Dr. Annaswamy has been elected to be a Fellow of the IEEE (2002) and IFAC (2017). She received a Distinguished Member Award and a Distinguished Lecturer Award from IEEE CSS in 2017. Dr. Annaswamy is actively involved in IFAC, IEEE, and IEEE CSS. She has served as General Chair of the American Control Conference (2008) as well as the 2nd IFAC Conference on Cyber- Physical & Human Systems (2018). She is Deputy Editor of the Elsevier publication Annual Reviews in Control (2016-present). She has been a member of IEEE Fellows Committee and the IEEE CSS Outreach Committee, and is the Chair of IEEE Smart Grid Meetings and Conferences. In IEEE CSS, she has served as Vice President of Conference Activities (2015-16) and Technical Activities (2017-18), and will serve as the President in 2020.

Warren Dixon received his Ph.D. in 2000 from the Department of Electrical and Computer Engineering from Clemson University. He worked as a research staff member and Eugene P. Wigner Fellow at Oak Ridge National Laboratory (ORNL) until 2004, when he joined the University of Florida in the Mechanical and Aerospace Engineering Department where he currently holds the Newton C. Ebaugh professorship. His main research interest has been the development and application of Lyapunov-based control techniques for uncertain nonlinear systems. His work has been recognized by a number of early career, best paper, and student mentoring awards. He is a Fellow of ASME and IEEE for his contributions to control of uncertain nonlinear systems.

K. Merve Dogan received a Bachelor of Science degree from Pamukkale University in 2014 and a Master of Science degree from Izmir Institute of Technology in 2016. She is currently a Doctor of Philosophy candidate in the Department of Mechanical Engineering at the University of South Florida, where she is also a member of the Laboratory for Autonomy, Control, Information, and Systems. Her research expertise includes adaptive control, robust control, distributed control, and convex optimization with applications to robotics, autonomous vehicles, multiagent systems, and human-machine interaction systems, where on these topics she has authored/co-authored over 40 published/submitted journal and conference papers.

Jonathan A. Muse received a Bachelor of Science degree from University of Alabama in 2005, a Master of Science degree from Georgia Institute of Technology in 2008, and a Doctor of Philosophy degree in Aerospace Engineering from Georgia Institute of Technology in 2010. He is currently a research engineer for the Aerospace Systems Directorate at the Air Force Research Laboratory (Wright-Patterson Air Force Base), where he is responsible for executing basic research related to hypersonic vehicle control and advanced methods in nonlinear control. He is the Co-PI for HIFiRE flight 6, an adaptive flight control experiment on a scramjet hypersonic vehicle currently being constructed.

Frank Lewis: Member, National Academy of Inventors. Fellow IEEE, Fellow IFAC, Fellow AAAS, Fellow U.K. Institute of Measurement & Control, PE Texas, U.K. Chartered Engineer. UTA Distinguished Scholar Professor, UTA Distinguished Teaching Professor, and Moncrief-O’Donnell Chair at The University of Texas at Arlington Research Institute. Founding Member Mediterranean Control Association. Qian Ren Thousand Talents Consulting Professor, Northeastern University, Shenyang, China. Ranked at position 81 worldwide, 59 in the USA, and 3 in Texas of all scientists in Computer Science and Electronics, by Guide2Research. Bachelor's Degree in Physics/EE and MSEE at Rice University, MS in Aeronautical Engineering at Univ. W. Florida, Ph.D. at Ga. Tech. He works in feedback control, reinforcement learning, intelligent systems, and distributed control systems. Author of 7 U.S. patents, 410 journal papers, 426 conference papers, 20 books, 48 chapters, and 12 journal special issues. He received the Fulbright Research Award, NSF Research Initiation Grant, ASEE Terman Award, Int. Neural Network Soc. Gabor Award 2009, U.K. Inst. Measurement & Control Honeywell Field Engineering Medal 2009. Received AACC Ragazzini Education Award 2018, IEEE Computational Intelligence Society Neural Networks Pioneer Award 2012 and AIAA Intelligent Systems Award 2016. IEEE Control Systems Society Distinguished Lecturer. Project 111 Professor at Northeastern University, China. Distinguished Foreign Scholar at Chongqing Univ. China. Received Outstanding Service Award from Dallas IEEE Section, selected as Engineer of the Year by Ft. Worth IEEE Section. Listed in Ft. Worth Business Press Top 200 Leaders in Manufacturing. Received the 2010 IEEE Region 5 Outstanding Engineering Educator Award and the 2010 UTA Graduate Dean’s Excellence in Doctoral Mentoring Award. Elected to UTA Academy of Distinguished Teachers 2012. Texas Regents Outstanding Teaching Award 2013. He served on the NAE Committee on Space Station in 1995.