Workshops

The ACC offers well-attended workshops addressing current and future topics in automatic control from experts from academia, national laboratories, and industry. In 2013, workshops will take place before the conference on Saturday June 15 and Sunday June 16. Please note that workshops are (a) subject to cancellation for lack of registrants, and (b) subject to capacity limits. Prospective attendees are therefore advised to register for their intended workshops early.

Conference registrants can sign up for workshops directly through the registration site. For workshop-only registration (i.e., registering for a workshop without registering for the conference proper), please contact the conference Registration Chair, Eric Frew with your name, e-mail address, and PaperPlaza PIN. See the Registration page for addtional details.

Advance registration for preconference workshops will be done through the registration site. Onsite registration at the conference venue will also be available.

Please register early to aid planning of the Workshop that you are interested in.

Saturday and Sunday Workshop June 15 and 16, 2013

Active Disturbance Rejection Control: Industrial Solutions to Process Uncertainties [Disturbance Rejection]
Saturday full day and Sunday half day (am), Room 16
Organizers: Zhiqiang Gao, Yi Huang, Enrico Canuto, Bao-Zhu Guo, and Hebertt Sira Ramirez

Saturday Workshops June 15, 2013

Robust State and Unknown Input Estimation: A Practical Guide to Design and Applications [Robust Estimation]
Full day, Room 2
Organizers: Stefen Hui, Stanislaw H. Zak

Gain Scheduled MPC: Design and Implementaion Using MATLAB [Gain Scheduled MPC]
Full day, Room 3
Organizer: Liuping Wang

Human Factors in Control [Human Factors]
Full day, Room 4
Organizers: James Brooks, Kit Siu, Anand Tharanathan

The ARIMA and VARIMA Time Series [Time Series]
Half day (pm), Room 5
Organizer: Ky Vu

Sunday Workshops June 16, 2013

Game-Theoretic Approach to Secure and Resilient Control Systems [Resilient Control]
Full day, Room 2
Organizers: Tamer Basar, Quanyan Zhu

ID of NL Parameter-Varying Systems: Theory and Applications [Nonlinear ID]
Full day, Room 3
Organizers: Wallace E. Larimore, Michael Buchholz

Decision Making Algorithms for Unmanned Vehicles [Unmanned Vehicles]
Full day, Room 4
Organizers: Sivakumar Rathinam, Krishnamoorthy Kalyanam, Swaroop Darbha

Emerging Applications in Systems and Control Theory for Neuroscience and Neural Medicine [Neuroscience and Neural Medicine]
Full day, Room 5
Organizers: ShiNung Ching, M. Brandon Westover, Sridevi Sarma

Health Management, Fault-tolerant Control and Cooperative Control of Unmanned Systems [Unmanned Systems] Cancelled.
Full day, Room 8
Organizers: Youmin Zhang, Camille Alain Rabbath, YangQuan Chen, Christopher Edwards, Cameron Fulford, Hugh Hong-Tao Liu, Vicenc Puig, Didier Theilliol, Khashayar Khorasani

Nonlinear Regression
Full day, Room 9
Organizer: R. Russell Rhinehart

Control of Power Inverters for the Smart Grid [Smart Grid]
Half (pm), Room 10/11
Organizer: Qing-Chang Zhong

Stochastic Models, Information Theory, and Lie Groups [Lie Groups]
Half (pm), Room 12
Organizer: Gregory Chirikjian

Workshop Descriptions

Saturday and Sunday Workshop, June 15 and 16, 2013

Active Disturbance Rejection Control: Industrial Solutions to Process Uncertainties
Organizers: Zhiqiang Gao, Yi Huang, Enrico Canuto, Bao-Zhu Guo, and Hebertt Sira Ramirez
Speakers: Hui Xie, Lei Guo, Rafal Madonski, Wen Tan, Wenchao Xue, Shen Zhao, Qing Zheng

Overview: This workshop focuses on the very nature of industrial control problems: obtaining consistent performance in processes full of uncertainties, the solution of which cannot be easily found in a model-based control theory. The essence of such problems is disturbance rejection, in a general sense, and that the question is how the disturbance is best mitigated, in view of new principles, methods, algorithms, and rigorous justifications. The solutions, to be truly industrial, must be easily scalable across product lines, or even industry sectors; and they must give what production engineers want: a control system that makes the difference small between the process variables and their desired values, and does so quickly, economically, and intuitively. The speakers in this workshop will show, interactively with the audience, that to achieve such a lofty goal, the disturbances must be rejected in an ACTIVE manner, meaning that the control system actively seeks and mitigates the causes of the output deviation, rather than just passively reacts to it. And this is the principle of Active Disturbance Rejection Control (ADRC).

Format: The workshop consists of three half-day sessions: 1) principles, methods and impacts, stretching from the steam engine era to present time, when ADRC is about to replace PID in a new line of TI chips; 2) technologies and applications, as ADRC is gaining a foothold in industry; 3) theoretical foundation, rigorous mathematical proofs, and the beginning of a theory of disturbance rejection. In each of the three sessions, the speakers will lead, but not dominate, the proceeding by giving short, concise presentations on the latest developments, followed by a forum of speakers-audience interactions on questions and concerns from participants, submitted either on-line or in person. Practical problems and examples are used throughout the discussions to help participants connect novel principles to their own experiences. Matlab simulations are shared with the participants to show the simplicity and effectiveness of ADRC. Benchmark problems from various domains of applications are used to demonstrate the novelty of ADRC.

Target Audience: practitioners looking for advanced control solutions that are simple to use and easy to tune; researchers looking for new ideas and the connection between theory and practice; students looking for better understanding of the foundation of controls and the essence of control engineering practice.

Workshop attendees are able to receive 1.2 CEU from the IEEE for attending this workshop.

Detailed information about the workshop can be found at this link.

Saturday Workshops, June 15, 2013

Robust State and Unknown Input Estimation: A Practical Guide to Design and Applications
Organizers: Stefen Hui (San Diego State University) and Stanisław H. Zak (Purdue University)

Observation and measurement play essential roles in achieving the control objective in a control scheme. Observed data are crucial to anticipate what might happen, or to explain what did happen. An observer is a deterministic dynamic system that can generate an estimate of the plant's state using that plant's input and output signals. Observers are utilized to augment or replace sensors in a control system. Since the early developments, which concentrated on observers for systems without uncertainties, observers for plants with both known and unknown inputs have been developed resulting in the so-called unknown input observer (UIO) architectures.

The aim of this workshop is to present in a tutorial fashion the theory and design of observers for uncertain systems. The intended audience includes practitioners who are primarily interested in applying modern control techniques, engineers who desire an introduction to the concepts and tools that can be used to design high performance systems capable of accommodating modeling uncertainties and system complexities, and faculty members, as well as graduate students, who wish to acquaint themselves with some of the more advanced techniques.

Robust state and unknown input observer architectures as well as their applications in diverse areas will be presented. In particular, the unknown input observer architectures will be applied to estimate stress in humans and to fault detection and isolation. These applications show the incredible potential of unknown input observers.

Presentations will be illustrated with simulations and implementations using MATLAB and Simulink. In addition there will be the workshop website available to the workshop participants. This interactive website will be using wiki software making it possible to share information using simple text formatting conventions. Workshop participants will be able to add comments on the spot.

For more information, see the website, [use login: guest password: acc2013]

Gain Scheduled MPC: Design and Implementation Using MATLAB
Organizer: Liuping Wang (RMIT University and NICTA, Victoria Research Laboratory, Australia)

Model Predictive Control (MPC) has a long history in the field of control engineering. It is one of the few areas that has received on-going interest from researchers in both the industrial and academic communities. Three major aspects of model predictive control make the design methodology attractive to both engineers and academics. The first aspect is the design formulation, which uses a completely multivariable system framework where the performance parameters of the multivariable control system are related to the engineering aspects of the system; hence, they can be understood and 'tuned' by engineers. The second aspect is the ability of method to handle both 'soft' constraints and hard constraints in a multivariable control framework. This is particularly attractive to industry where tight profit margins and limits on the process operation are inevitably present. The third aspect is the ability to perform process on-line optimization.

The model predictive control systems are designed using linear models unless a nonlinear model is explicitly stated. Nonlinear model predictive control is conceptually similar to its linear counterpart except that nonlinear models are deployed for the prediction and optimization. However, because of its computational intensity and complexity, the nonlinear predictive control systems are not as widely applied as its linear counterpart. Instead, the gain scheduled control system techniques have found success in the area of predictive control of nonlinear plants. This one-day short-course will show the four steps involved in the design of a gain scheduled predictive controller: (i) linearization of a nonlinear plant model according to operating conditions; (ii) the design of linear predictive controllers using the family of linear models; (iii) gain scheduled predictive control law that will optimize a multiple model objective function with constraints, which will also ensure smooth transitions (i.e. bumpless transfer) between the predictive controllers; (iv) simulation and experimental validation of the gain scheduled predictive control system with constraints using MATLAB® and Simulink® as a platform. The course, based on a predictive control book (second edition, in preparation) by the speaker, is suitable for engineers, students and researchers who wish to gain basic knowledge about gain scheduled model predictive control of nonlinear plant, as well as understand how to perform real time simulation and implementation using MATLAB and Simulink tools.

For additional information, see PDF file.

Human Factors in Control
Organizers: James Brooks (Real-Time Optimization and Controls Lab, GE Global Research), Kit Siu (Model Based Controls Lab, GE Global Research), Anand Tharanathan (Human Factors Center of Excellence, Honeywell Advanced Technology Labs)

The purpose of this workshop is to provide a forum which introduces the broader controls community to theoretical frameworks of human performance (from human factors, applied psychology, and cognitive science) which are relevant to the interaction between humans and the systems designed around them as well as highlight relevant, active application and research areas. The selected talks and speakers are a blend of theory from academia as well as current applications in industry.

All control systems are designed to meet the needs perceived by humans as well as interact with humans on some level. As control systems become more inclusive of complete systems, human factors will become an increasingly important consideration in the design and analysis of control strategies.

Introductory topics in human performance and automation interaction presented will include situation awareness, decision making, and interface design. Applied talks include a presentations by Honeywell and General Electric regarding implications for process control display designs and user experience for next generation human-machine interfaces. Speakers include Mica Endsley (SA Technologies and current President of the Human Factors and Ergonomics Society), Stephanie Guerlain (Univ. Virginia), Ling Rothrock (Penn State), Anand Tharanathan (Honeywell Advanced Technology Labs), and So Young Kim (GE Global Research).

This workshop will be beneficial for systems engineers (in both academia and industry) working on, or interested in, control systems and/or complex systems with significant human interaction and perception requirements as well as graduate students in relevant areas. These areas include human performance and intent modeling, human supervisory control, interface design, manual control, human-in-the-loop systems, and perceptions of optimality by operators.

Participants can expect to:

• a) be introduced to key theoretical concepts in human performance
• b) methodological approaches to studying human-machine systems
• c) engage with others interested in human factors as it relates to control system and interface design\

For additional information, see PDF file.

The ARIMA and VARIMA Time Series
Organizer: Dr. Ky M. Vu (AuLac Technologies Inc., Canada)

This workshop is an introduction to the discrete time series: the scalar ARIMA and its vector counterpart VARIMA time series. This workshop is a result of more than twenty years of interest in, and research of, the presenter on the topic of the ARIMA time series. The ARIMA and its vector counterpart VARIMA time series have vast applications in many different fields of science and technologies. In control theory, it has application in designing a stochastic controller.

The workshop will cover the following topics:

1. Basic probability and statistics.
2. Deterministic (periodic) and the Stochastic time series (ARIMA and VARIMA).
3. Time (Moment) and frequency (spectra) statistics.
4. Modeling, forecast and controller/filter designs.
5. Future research on the ARIMA time series.

After attending the workshop, an attendee will be able to do the following:

1. Designing a stochastic controller.
2. Designing or doing research on digital filters.
3. Analyzing or making forecast of discrete time data.

The workshop notes will be extracted from the textbook "The ARIMA and VARIMA Time series: Their Modelings, Analyses and Applications." Demonstrations will be running MATLAB m files from the toolbox V-ARIMA Time Series. Huge discounts on the textbook and toolbox are given to attendees. More information on the textbook and toolbox are available from the website www.aulactechnologies.com. Each attendee should bring with him or her a notebook (laptop) with MATLAB software, or a clone like OCTAVE, installed on it.

For additional information, see PDF file.

Sunday Workshops, June 16, 2013

Game-Theoretic Approach to Secure and Resilient Control Systems
Organizers: Tamer Basar, Quanyan Zhu (University of Illinois at Urbana-Champaign)

Intelligent control systems are pervasive in modern critical infrastructures such as power grids, transportation systems, healthcare informatics, etc. The system level integration of cyber and physical components has made control systems vulnerable to malicious events and cyber attacks. Distinct from traditional measures of robustness and reliability, security and resilience are important system attributes of control systems that are often subject to exogenous disturbances and exposed to unexpected adversarial attacks or events. Hence, there is need for fundamental understanding of these concepts in the context of control systems, which can lead to a new set of computational and system design tools. The proposed full-day workshop aims to address security and resilience aspects of control systems using decision and game-theoretic methods. The workshop will consist of topics covering different aspects of resilient control systems including secure communications, network security, control system management, resilient control design and human factors. The workshop will also draw examples from communication networks and smart energy systems to illustrate various applications of game theory.

ID of NL Parameter-Varying Systems: Theory and Applications
Organizers: Wallace E. Larimore (Adaptics, Inc.), Michael Buchholz (Institute of Measurement, Control and Microtechnology, Ulm University)

In this workshop, major recent advances in system identification for parameter-varying and nonlinear systems using linear time-invariant (LTI) subspace methods are developed. Other such system identification methods that are currently available include subspace methods where the required computation grows exponentially with the number of system inputs, outputs, and states; or they involve iterative nonlinear parameter optimization that can encounter convergence difficulties.

The workshop presents a first principles statistical approach using the fundamental canonical variate analysis (CVA) method for subspace identification of linear time-invariant (LTI) systems, with detailed extensions to linear parameter-varying (LPV) and nonlinear systems. The LTI case includes basic concepts of reduced rank modeling of ill-conditioned data to obtain the most appropriate statistical model structure and order using optimal maximum likelihood methods. The fundamental statistical approach gives expressions of the multistep-ahead likelihood function for subspace identification of LTI systems. This leads to direct estimation of parameters using singular value decomposition type methods that avoid iterative nonlinear parameter optimization. The result is statistically optimal maximum likelihood parameter estimates and likelihood ratio tests of hypotheses. The parameter estimates have optimal Cramer-Rao lower bound accuracy, and the likelihood ratio hypothesis tests on model structure, model change, and process faults produce optimal decisions. Comparisons made between system identification methods including subspace, prediction error, and maximum likelihood show considerably less computation and higher accuracy.

The LTI subspace methods are extended to LPV systems that are in the LTI form where the constant LTI parameters are multiplied by parameter-varying scheduling functions depending on the system operating point. For example, this allows for the identification of constant underlying structural stiffness parameters while wing flutter dynamics vary with scheduling functions of speed and altitude operating point variables. This is further extended to Quasi-LPV systems where the scheduling functions may be functions of the inputs and/or outputs of the system. Quasi-LPV systems include bilinear and general polynomial systems that are universal approximators. Identified models of such nonlinear parameter-varying systems provide a starting point for gain scheduling and global control design. Applications are discussed to monitoring and fault detection in closed-loop chemical processes, identification of vibrating structures under feedback, online adaptive control of aircraft wing flutter, identification of the chaotic Lorenz attractor, and identification and monitoring of nonlinear automotive engines.

Schedule:

• Start time: 8:00
• Morning coffee break: 9:45-10:00
• Lunch break: 11:45 – 12:45
• Afternoon coffee break: 2:45-3:00
• End time: 5:00.

Decision Making Algorithms for Unmanned Vehicles
Organizers: Sivakumar Rathinam, Krishnamoorthy Kalyanam, Swaroop Darbha
Speakers from Air Force Research Lab: Dr. Meir Pachter, Dr. Krishnamoorthy Kalyanam
Speaker from University of Florida: Dr. Pramod Khargonekhar
Speakers from Texas A & M University: Dr. Sivakumar Rathinam and Dr. Swaroop Darbha

This workshop is focused on disseminating recent advances in the area of decision making algorithms for Unmanned Vehicles (UVs). Small autonomous UVs are seen as ideal platforms for many military applications such as monitoring targets, mapping a given area, aerial/road surveillance and civil applications such as search and rescue missions, reconnaissance, fire monitoring etc. The main advantage of using these vehicles is that they can be used in situations where a manned mission is dangerous and/or is not possible. Even though there are several advantages in using small UVs, these vehicles have resource constraints due to their limited computational and sensing capabilities. To successfully realize a mission involving UVs, one has to optimally use the available resources often in the presence of uncertainty, motion and sensing constraints, and possibly adversarial action. This workshop brings together practitioners from the academia and the Air Force Research Lab (AFRL), to provide a complete picture of the challenges behind UVs mission planning, specification, scheduling and real-time optimization. The talks in this workshop will discuss open problems, touch upon various issues in the area of UV decision making, provide practical solutions for some of the underlying problems and also illustrate the theoretical background behind the techniques. In particular, the talks will focus on the following topics:

1. An overview of UV decision making problems under uncertainty
2. Constrained motion planning algorithms for a collection of UVs
3. Approximation algorithms for UVs with resource and motion constraints
4. Role of information in dynamic games involving UVs and adversaries
5. Approximate linear programming methods for stochastic dynamic programs and its application to UVs
6. Intruder isolation and capture on a road network using partial and delayed information

For additional information, see PDF file. Also, see the website.

Emerging Applications in Systems and Control Theory for Neuroscience and Neural Medicine
Organizers: ShiNung Ching, M. Brandon Westover, Sridevi Sarma

The past several years have seen rapid advances in the development of medical technologies that hold exciting promise as new therapeutic options in neurological disease and new neuroscience tools for understanding brain function. Achieving this promise will require the development of methods to robustly model, estimate and regulate the dynamics of highly complex neuronal circuits and systems. Systems and control theory will, thus, be a critical ingredient in the successful realization of these new medical and scientific advances.

This workshop will offer a survey of emerging problems at the boundary between systems and control engineering, neuroscience and neural medicine. The topics will span multiple scales of control: (1) closed loop control of large scale brain activity using neural pharmacology, (2) estimation and control of brain and neuronal dynamics in disease using brain stimulation technology and (3) control of small scale activity in local neuronal circuits and individual neurons using experimental neuroscience methods. Emphasis will be placed on both recent theoretical developments – for instance, system identification and controllability in neuronal networks – as well as on real-world clinical constraints in applications such as deep brain stimulation for the treatment of neurological disease, or optical stimulation in experimental neuroscience.

For more information, see the website. The schedule is available as a PDF file.

Nonlinear Regression
Organizer: R. Russell Rhinehart (Oklahoma State University)

This workshop will be a practical guide for nonlinear regression modeling. Although theoretical analysis behind techniques will be revealed, the takeaway will be your ability to:

• Define the regression objective,
• Choose an optimization approach,
• Design experiments for data generation,
• Validate models with data and statistical techniques, and
• Select an appropriate model order (complexity).

The course is intended for engineering employees, students, or faculty seeking to match models to data.

A pre-textbook monograph of approximately 200-pages will be used as the workshop notes. Exercises and code can be implemented in any environment, but Excel/VBA will used as in-workshop examples and exercises. Participants are invited to bring a laptop with Excel version 2010 or higher to process examples during the workshop.

Data-based models are often central for model-based control, forecasting, training simulators, analysis and diagnosis, mechanism validation, and supervisory optimization. For many of these applications nonlinear models are preferred in order to capture the process/device behavior. Regression is the procedure of fitting models to data, and nonlinear regression means the adjustable model coefficients do not appear linearly within the model.

Since processes/devices are neither ideal nor stationary, model coefficients need to be adjusted periodically to make the models fit experimental data. Even for seemingly linear models, a variable delay introduces a nonlinear model coefficient.

Workshop topics will include equation structures, optimization of model coefficient values in the presence of constraints and local traps, choosing optimization stopping criteria based on model properties, data pre-processing and post-processing, data-based model validation, discrimination between models, design of experiments that support validation outcomes, propagation of uncertainty, and model utility evaluation.

This is not the standard linear regression approach to develop response surface model structures, and classic experimental designs. This workshop will focus on techniques for nonlinear regression.

AM – Session 1 8:00-9:45am – Introductory concepts, Model types and equation structures, Regression target (least squares or maximum likelihood), Constraints, Traps

AM – Session 2 10:00-11:45am – The Distortion of linearizing transforms, Optimization algorithms, Multiple optima Propagation of uncertainty

Lunch – 11:45am-1:00pm

PM – Session 3 1:00-2:45pm – Regression stopping criterion based on model properties, Undesired model characteristics, Data pre- and post-processing

PM – Session 4 3:00-5:00pm – Data-based model validation and model discrimination, Design of experiments to support validation outcomes, Evaluation of model utility (vs. perfection)

Control of Power Inverters for the Smart Grid
Organizer: Qing-Chang Zhong (The University of Sheffield)

Energy and sustainability are now on the top agenda of many governments. Smart grids have become one of the main enablers to address energy and sustainability issues. Renewable energy, distributed generation, hybrid electrical vehicles, more-electric aircraft, all-electric ships, smart grids etc will become more and more popular. Arguably, the integration of renewable and distributed energy sources, energy storage and demand-side resources into smart grids, often via inverters, is the largest "new frontier" for smart grid advancements. Control and power electronics are the two key enabling technologies for this. Power electronics is becoming part of the grid and control is where the "smart" is from. Together with the power systems infrastructure, they form the backbone for smart grids. An inverter, which converts DC electricity from sources such as batteries, solar panels, or fuel cells to AC electricity, are the interface to integrate renewable energy and distributed generation into smart grids. Hence, how to control inverters so that renewable energy sources can be connected to the grid in a smart and friendly way is very important for the modeling, operation and management of smart grids. Inverters are also widely used in uninterruptible power supplies (UPS), induction heating, high-voltage DC (HVDC) transmission, variable-frequency drives, electric vehicle (EV) drives, air conditioning, vehicle-to-grid (V2G) etc and, hence, have become a common key device for many energy-related applications. There are several important control problems associated with inverters. For example, how to make sure that the total harmonic distortion (THD) of the inverter voltage and current remains within certain range when the loads are nonlinear and the grid voltage, if present, is distorted; how to make sure that a balanced neutral line is provided for applications where a neutral line is needed, e.g. when three-phase loads are not balanced; how to make sure that inverters can be operated in grid-connected mode or standalone mode and how to minimize the transient dynamics when the operation mode is changed; how to synchronize inverters with the grid so that they can be connected to the grid when needed; how to share loads proportionally according to their power ratings when inverters are operated in parallel; how to connect inverters to the grid in a grid-friendly manner so that the impact on the grid is minimized etc. This workshop will focus on presenting advanced control strategies, accompanied with experimental results, to address these problems systematically; based on the research monograph entitled Control of Power Inverters in Renewable Energy and Smart Grid Integration recently published by Wiley-IEEE Press (the book is available from IEEExplore and Wiley Online Library).

For additional information, see PDF file.

Stochastic Models, Information Theory, and Lie Groups
Organizer: Gregory Chirikjian, Johns Hopkins University

This tutorial reviews stochastic models in robot motion planning, state estimation, and control, and explains how concepts from both information theory and the theory of Lie groups are naturally intertwined with these models. Unlike most treatments of continuous time stochastic processes, which are either highly formalized or focusing on one-dimensional problems, the review provided here will cover physically motivated multi-dimensional problems in an accessible way.

Topics such as the relationship between nonholonomic constraints, white noise forcing, and the parametric distributions that result, will be discussed. The inter-conversion of Ito and Stratonovich stochastic differential equations will be explained. Some elementary Lie group theory will be used.

Application areas will include orientation and pose estimation, simultaneous localization and mapping, and steering flexible needles for minimally invasive surgical applications. As time permits other topics will be discusses including stochastic modeling of biological macromolecules from a systems-theoretic perspective.

This presentation will be based on material from the organizer's new two-volume book of the same title.

For further information on the ACC 2013 workshops please contact the conference Workshops Chair, Linda Bushnell.