Main

Symposium on Modeling and Optimization

Webinars are held about every two-weeks. These seminars consist of applications and tutorials in mathematical modeling, estimation, and optimization.

Topic Date Time Presenter Affiliation
Optimization of Olefins Production Oct 17, 2013 4 PM MDT Mark Darby CMiD Solutions
Process Control in the Upstream Industry TBD 4 PM MDT Michael Nikolaou University of Houston
UAVs for Infrastructure Monitoring Oct 31, 2013 4 PM MDT Kevin Franke BYU Civil Engineering
Advanced Deep Water Monitoring Systems Nov 7, 2013 4 PM MST David Brower Astro Technology, Inc.

To participate in an upcoming presentation, click the link above to join the WebEx session. There is no charge to join the webinar and Windows, Linux, iOS, and mobile platforms are supported. Users can participate via audio or video conference through a telephone or internet connection. Presentations are recorded and posted to YouTube as is the case with many of the past sessions shown below.

Past Presentations

Topic Presentation Date Presenter Description
Stimulation Optimization of Unconventional Resources Presentation (55 min) Jan 18, 2013 Marisela Sanchez-Nagel Improved Numerical Modeling from First Physics Emerging numerical techniques are available that can address hydraulic fracturing in naturally fractured formations, such as shale gas, and that are based upon the proper first physics of the mechanical and flow behavior of fractured formations. These techniques will allow the industry to gain greater insight and understanding of how natural fracture geometry and intensity, fracture hydromechanical properties, and stress field affect hydraulic fracturing performance in the presence of different operational parameters.
Efficient Surrogate Model Generation Presentation (43 min) Nov 13, 2012 John Eason and Selen Cremaschi
Univ of Tulsa 		Surrogate models, by constructing simpler functional approximations of complex and computationally expensive models, allow the use of traditional optimization algorithms for otherwise intractable problems. This talk discusses three possible algorithms for efficient surrogate model building.
Fast NMPC and MHE tools Presentation (57 min) Nov. 6, 2012 Milan Vukov and Moritz Diehl, KU Leuven Fast estimation and control techniques are critical for real-time applications. Researchers at KU Leuven will share cutting-edge methods to apply Nonlinear Model Predictive Control (NMPC) and Moving Horizon Estimation (MHE) for real-time applications.
Proactive Energy Management Strategies for Buildings Presentation (34 min) Oct. 23, 2012 Cara R. Touretzky
Michael Baldea
UT Austin 		Buildings are dynamic systems defined by fluctuations in weather, occupancy and energy prices. A nonlinear model-based optimal energy management strategy demonstrates significant energy savings compared to setpoint tracking strategies.
Mixed Integer Nonlinear Programming Presentation #1 (28 min)

Slides #1

Presentation #2 (14 min)

Slides #2 		Oct. 2, 2012 Wesley Cole, UT Austin

Jeremy Castagno, BYU 		Many design, engineering, and scientific applications include a mixture of continuous and discrete decision variables and nonlinear relationships. Mixed Integer Nonlinear Programming (MINLP) solvers are reviewed and compared for application to a scheduling problem for UT Austin central campus cooling.
Generalized Disjunctive Programming Presentation (34 min) Sept. 18, 2012 Juan Ruiz, Visual MESA Mixed-Integer Programming (MIP) provides a framework for mathematically modeling many optimization problems that involve discrete and continuous variables. Even though there has been a significant increase in the use of these models, a question of how to develop the best model that will lead to the most efficient solution method remains to be answered. In this presentation we review the latest developments in Generalized Disjunctive Programming (GDP), a modeling framework that aims at tackling the above question.
MPECs in Integrated Energy Storage, Forecasting, and Optimization
Intro_1
(2.5 min)

Intro_2
(3 min)

Presentation (30 min) 		Sept. 4, 2012 Kody Powell, UT Austin Optimizing the University of Texas at Austin utilities (heat, cooling, and power) plant requires the use of conditional arguments that change the system dynamics. New formulations of Mathematical Programs with Equilibrium Constraints (MPECs) have improved convergence properties to enable more realistic dynamic simulations.
Mars Curiosity Introduction (2.5 min)

Presentation (59 min)

Slides 		Aug. 30, 2012 Todd Barber, NASA JPL The Curiosity robot is equipped with a nuclear-powered lab capable of vaporizing rocks and ingesting soil, measuring habitability, and potentially paving the way for human exploration. Todd will review interesting and unusual experiences in his career as a NASA engineer. He will present audio-visual highlights and recent findings of the Mars Rover Missions. (Spirit, Opportunity, and Curiosity).
MHE for UAVs Introduction (3 min)

Presentation (28 min) 		Aug. 29, 2012 Liang Sun, BYU Moving Horizon Estimation (MHE) is used to update a mathematical model during real-time flight tests. Results from the flight tests demonstrate that large-scale models can be applied in real-time applications for estimation and control.
DAE Solution for Separation Processes Presentation
Webinar Video 		July 17, 2012 David Harney The equations that simulate separation processes naturally involve higher index Differential and Algebraic Equations (DAEs). Other popular techniques for DAE simulation can often only handle up to index-1 DAEs. New techniques for large-scale DAEs allow efficient solution for bifurcation analysis, simulation of complex systems, and stability analysis.
Predicting Protein Structures Presentation May 22, 2012 Jose Mojica
Reza Asgharzadeh
Thomas Knotts
Josh Price 		Prediction of protein structures is a rapidly advancing area of research that relies on parallel computation, advanced optimization techniques, and experimental validation. A genetic algorithm approach is detailed as one of the optimization techniques to solve the mixed-integer nonlinear programming problems. Methods to align experimental results with simulation are discussed.
Optimization of Design Under Uncertainty Presentation May 8, 2012 Casey Abbott Process design has a number of unknown or uncertain parameters that influence the optimal design, desired operational conditions, and project profitability. Optimization of design maximizes the expected NPV (Net Present Value) considering uncertain design parameters. Download example application
SBML Models in APM Presentation
Slide Notes 		April 23, 2012 David Grigsby The Systems Biology Markup Language (SBML) is standard format for expressing dynamic models in computational biology. This presentation will provide a tutorial on using large-scale SBML models for dynamic simulation, parameter estimation, and optimization. The tutorial will include instruction on using the APM conversion utility and solving SBML example problems.
Fast Model Predictive Control Presentation April 3, 2012 Trevor Slade and Reza Asgharzadeh Traditionally, Model Predictive Control (MPC) is executed every 15-60 seconds. Recent developments have explored storage and retrieval to increase the speed. This work is an implementation of Fast Model Predictive Control (1-5 Hz) without storage and retrieval for a four tank process.
Optimal Boiler Control Presentation March 20, 2012 Jose Mojica Load following in power generation is a recent opportunity as time of day pricing and cogeneration are adopted in refining, chemical, and power plants. This work is to improve the load following capabilities of natural gas and coal-fired boilers. Download example application
Optimal Identification for Model Predictive Control Presentation March 13, 2012 Mark Darby To identify models for MPC applications in chemical plants and refineries, the process is perturbed through increasingly automated step testing. Mark will give a tutorial overview of how to apply practical control engineering and design experiments for model development of dynamic systems.
Thermal Oxidizer Boundary Value Management / InBound Presentation 1
Presentation 2 		March 6, 2012 Reza Asgharzadeh, Chris Lyden, and Jim Huff Thermal oxidizers are used in chemical plants and refineries to combust waste streams with low concentrations of reactants. The design and operation of the thermal oxidizer is of crucial importance for safety, environmental, and economic reasons.
Nonlinear Programming with APM MATLAB Presentation February 21, 2012 John Hedengren The APM interface extends MATLAB to be used for parameter estimation, nonlinear control, and optimization. Participants can download a few example applications to run prior to the meeting that will be covered as part of the tutorial.
Nonlinear Programming with APM Python Presentation February 7, 2012 John Hedengren The APM interface extends Python to be used a variety of optimization applications. Dynamic optimization programming applications are demonstrated with Python.
Solid Oxide Fuel Cell Modeling and Control Presentation January 24, 2012 Lee Jacobsen Solid Oxide Fuel Cells (SOFCs) can be damaged by load changes that shorten life through delamination and micro-cracking. Application of SOFCs in load following applications is discussed to improve lifecycle costs.
Aerial Vehicle Control for Drogue Following Presentation January 17, 2012 Solomon Sun Path optimization of an Unmanned Aerial Vehicle (UAV) becomes more challenging with select constraints. Application of a UAV application is discussed from the MAGICC Lab group at BYU.
Dynamic Energy Storage Presentation December 13, 2011 Kody Powell, UT Austin Dynamic thermal energy storage with weather forecasting has the potential to improve solar energy capture by up to 64%. Results from a recent study are discussed to demonstrate the benefits of dynamic optimization.
Virus and Biological Modeling Presentation November 29, 2011 Casey Abbott Large scale biological models are important to gain an understanding of complex pathways to improve treatments such as anti-viral application. Simulated clinical trial data is aligned to an available model of HIV infection.
Friction Stir Welding Presentation November 15, 2011 Dustin Marshall Friction Stir Welding (FSW) temperature control is important to achieve uniform and consistent weld properties. This application is using a PDAE model for estimation and control. The model validation, control, and experimental data testing are discussed.