Hoagg, Jesse B

Owned by UKY CCS RCD
Jun 09, 2024

Research information updated - 02/12/2018

A Control-Systems Approach to Understanding Human Learning


Description:
Humans learn to control a wide range of dynamic systems; however, the strategies used by humans to control these systems are not well understood. Our research addresses fundamental questions of human-learning science: How do humans learn to control dynamic systems? What control strategies do humans learn? What characteristics make a system difficult for a human to learn to control?
In this research, we employ a system-theoretic approach to studying human learning. We conduct experiments in which human subjects repeatedly interact with dynamic systems, and through these interactions, the subjects learn to control the systems. We developed a new subsystem identification (SSID) approach, which uses experimental input-output data to construct mathematical models of the subjects’ control be- havior. These identified models provide new insights into human learning and the characteristics that make dynamic systems difficult for humans to control.

Students:

  1. 1) Xingye Zhang, Ph.D. in Mechanical Engineering, University of Kentucky, 2015
  2. 2) Seyyed Alireza Seyyed Mousavi, Ph.D. Student in Mechanical Engineering, University of Kentucky
  3. 3) Shaoqian Wang, Ph.D. Student in Mechanical Engineering, University of Kentucky
  4. 4) Sajad Koushkbaghi, Ph.D. Student in Mechanical Engineering, University of Kentucky
  5. 5) Amelia Sheffler, M.S. Student in Mechanical Engineering, University of Kentucky


Publications:

  1. X. Zhang, S. Wang, J. B. Hoagg, and T. M. Seigler, “The roles of feedback and feedforward as humans learn to control unknown dynamic systems,” IEEE Transactions on Cybernetics, vol. 48, no. 2, pp. 543–555, 2018. DOI: 10.1109/TCYB.2016.2646483
  2. X. Zhang and J. B. Hoagg, “Subsystem identification of multivariable feedforward and feedback systems,” Automatica, vol. 72, pp. 131–137, 2016. DOI: 10.1016/j.automatica.2016.05.027
  3. X. Zhang and J. B. Hoagg, “Frequency-domain subsystem identification with application to modeling control strategies used by humans,” Systems & Control Letters, vol. 87, pp. 36–46, 2016. DOI: 10.1016/ j.sysconle.2015.10.009
  4. F. Matveeva, S. A. Seyyed Mousavi, X. Zhang, T. M. Seigler, and J. B. Hoagg. “On the effects of changing reference command as humans learn to control dynamic systems,” Proc. Conf. Dec. Contr., pp. 1211–1216, Las Vegas, NV, December 2016. DOI: 10.1109/CDC.2016.7798431
  5. S. A. Seyyed Mousavi, X. Zhang, T. M. Seigler, and J. B. Hoagg. “Characteristics that make dynamic systems difficult for a human to control,” Proc. Amer. Contr. Conf., pp. 4391–4396, Boston, MA, July 2016. DOI: 10.1109/ACC.2016.7525613
  6. X. Zhang, T. M. Seigler, and J. B. Hoagg, “Modeling the control strategies that humans use to control nonminimum-phase systems,” Proc. Amer. Contr. Conf., pp. 471–476, Chicago, IL, July 2015. DOI: 10.1109/ACC.2015.7170780
  7. X. Zhang, S. Wang, T. M. Seigler, and J. B. Hoagg, “Frequency-domain observations on how humans learn to control an unknown dynamic system,” Proc. Amer. Contr. Conf., pp. 1143–1148, Chicago, IL, July 2015. DOI: 10.1109/ACC.2015.7170887
  8. X. Zhang, S. Wang, T. M. Seigler, and J. B. Hoagg, “A subsystem identification technique for modeling control strategies used by humans,” Proc. Amer. Contr. Conf., pp. 2827–2832, Portland, OR, June 2014. DOI: 10.1109/ACC.2014.6859211


Grants:

This work have been supported in part by the National Science Foundation and the Kentucky Engineering and Science Foundation through the following awards:

  1. A Control-Systems Approach to Understanding Human Learning J. B. Hoagg (PI); T. M. Seigler (co-PI) National Science Foundation (CMMI-1405257) $249,457 August 1, 2014 to July 31, 2018


  1. Assistive Learning for Human-Machine-Interaction Systems J. #B. Hoagg (PI); T. M. Seigler (co-PI) Kentucky Science and Engineering Foundation (KSEF-3453-RDE-018) $30,000 July 1, 2015 to December 31, 2016


  1. A Dynamic Systems Approach to Understanding Human Learning T. M. Seigler (PI); J. B. Hoagg (co-PI) Kentucky Science and Engineering Foundation (KSEF-148-502-12-288) $49,999 June 1, 2012 to December 31, 2013


A Dynamic Systems Approach to Understanding Human Learning

T. M. Seigler, Jesse B. Hoagg, Shaoqian Wang, and Xingye Zhang.
August 08, 2013
Description:
We use the DLX cluster for a project titled “A Dynamic Systems Approach to Understanding Human Learning”. Our project aims to modeling human learning by using techniques from dynamic systems and control theory. To date, we have conducted experiments aimed at studying human learning and human motor control. These human-learning experiments involve human test subjects, and we use custom system identification software to identify (or estimate) dynamic models that approximate the behavior of the subjects. Our goal is to use physical data to model how humans interact with their physical environment in a variety of scenarios. Our custom system identification approach is a multistep nonlinear optimization, which is computational intensive. Thus, our system identification approach is only tractable with parallel computing resources (e.g., DLX).

Computation Methods

The computational method is a multistep nonlinear optimization, which relies on iterative convex optimizations (i.e., linear least square). The algorithms for this project have been developed by the research groups of Dr. Seigler and Dr. Hoagg (Department of Mechanical Engineering).

Software

Matlab is used for this project.

Collaborators

The University of Kentucky personnel working on this project are:
Dr. Thomas M Seigler, Associate Professor, ME
Dr. Jesse B Hoagg, Assistant Professor, ME

Students

• Shaoqian Wang, PhD student, Department of Mechanical Engineering
• Xingye Zhang, PhD student, Department of Mechanical Engineering
• Seyyedalireza Seyyedmousavi, Graduate, Department of Mechanical Engineering

• Pedram Rabiee, UnderGrad, Department of Mechanical Engineering, 01/25/2019 

  • This project does not involve collaborators outside of the University of Kentucky.

Grants

This project is funded by the Kentucky Science and Engineering Foundation. Proposals based on this work have been submitted to the National Science Foundation.

Hoagg, Jesse B NNX10AL96H Curriculum Development: An Algebraic Approach to Linear-Quadratic Control National Aeronautics and Space Administration 5/1/2013 - 7/31/2014 SCOPE
Hoagg, Jesse B CMMI-1405257 A Control-Systems Approach to Understanding Human Learning National Science Foundation 8/1/2014 - 7/31/2017 $249,457
Hoagg, Jesse B NNX10AL96H Second Annual KIAE Kentucky Wing Design Competition National Aeronautics and Space Administration 1/1/2012 - 12/31/2014 SCOPE
Hoagg, Jesse B NNX10AL96H Discrete-Time Linear-Quadratic Control: An Algebraic Approach National Aeronautics and Space Administration 1/1/2014 - 12/31/2014 SCOPE

Center for Computational Sciences