Global Optimization of Unconventional Formulations for Sustainable Energy Systems
Event details
| Date | 10.06.2016 |
| Hour | 10:15 › 11:15 |
| Speaker | Prof. Alexander Mitsos, RWTH Aachen University, Germany |
| Location | |
| Category | Conferences - Seminars |
The presentation gives an overview of the group's recent and current work on energy systems and describes some highlights in more detail. The focus is on the optimal design of renewable energy systems, in particular concentrated solar thermal, geothermal and biomass conversion.
We discuss optimal layout, optimal design and optimal operation over time scales (hourly, daily, seasonal). We also discuss hybrid systems consisting of more than one energy source with the goal of synergetically combining sources with different characteristics.
These systems have characteristics that are different from conventional energy systems, namely the combination of turbomachinery and chemical processes, pronounced uncertainty and time-variability. As a result, these systems need to be addressed using systematic methods from process systems engineering and at the same time can be utilized to identify and overcome limits in these methodologies. In the last decade, algorithms for deterministic global optimization of mixed-integer nonlinear programs (MINLP) have undergone major changes rendering the solution of challenging optimization problems possible. We utilize these advantages to optimize several renewable energy systems using state-of-the-art codes. However, we also demonstrate that the optimization of several systems is not possible using these codes due to mainly two reasons, namely scaling of computational complexity with size of problems and the fact that natural formulations for many problems do not follow the traditional MINLP.
To address these problems we develop theory for the scaling of algorithms and algorithms for unconventional formulations and utilize them for energy systems. In particular we discuss embedded/hierarchical programs (bilevel, semi-infinite, differential-algebraic systems with optimization embedded) and how these problems can be solved. We finally briefly discuss open challenges in optimization theory and algorithms and in energy systems.
Bio: Alexander Mitsos is a Full Professor (W3) in RWTH Aachen University, and the Director of the Laboratory for Process Systems Engineering (AVT.SVT). His research focuses on optimization (design and operation) of energy and chemical systems and development of enabling numerical algorithms.
Educational Background
Academic Appointments
Professional Development
We discuss optimal layout, optimal design and optimal operation over time scales (hourly, daily, seasonal). We also discuss hybrid systems consisting of more than one energy source with the goal of synergetically combining sources with different characteristics.
These systems have characteristics that are different from conventional energy systems, namely the combination of turbomachinery and chemical processes, pronounced uncertainty and time-variability. As a result, these systems need to be addressed using systematic methods from process systems engineering and at the same time can be utilized to identify and overcome limits in these methodologies. In the last decade, algorithms for deterministic global optimization of mixed-integer nonlinear programs (MINLP) have undergone major changes rendering the solution of challenging optimization problems possible. We utilize these advantages to optimize several renewable energy systems using state-of-the-art codes. However, we also demonstrate that the optimization of several systems is not possible using these codes due to mainly two reasons, namely scaling of computational complexity with size of problems and the fact that natural formulations for many problems do not follow the traditional MINLP.
To address these problems we develop theory for the scaling of algorithms and algorithms for unconventional formulations and utilize them for energy systems. In particular we discuss embedded/hierarchical programs (bilevel, semi-infinite, differential-algebraic systems with optimization embedded) and how these problems can be solved. We finally briefly discuss open challenges in optimization theory and algorithms and in energy systems.
Bio: Alexander Mitsos is a Full Professor (W3) in RWTH Aachen University, and the Director of the Laboratory for Process Systems Engineering (AVT.SVT). His research focuses on optimization (design and operation) of energy and chemical systems and development of enabling numerical algorithms.
Educational Background
- Ph.D. in Chemical Engineering, June 2006, MIT
- Dipl.-Ing. in Chemical Engineering, March 1999, Karlsruhe University, Germany
Academic Appointments
- Sep 2012-, Professor for Process Systems Engineering, Aachener Verfahrenstechnik - AVT der RWTH Aachen, Deutschland
- Jan 2009-Aug 2012, Rockwell International Assistant Professor, Department of Mechanical Engineering, MIT
- Sep-Dec 2008, Visiting Scientist, Department of Mechanical Engineering, MIT
- Jan-Dec 2008, Junior Research Group Leader, AICES, RWTH Aachen
Professional Development
- Sep 2006-Dec 2007, Senior Engineer, RES Group Inc., Cambridge MA
- Mar 2001-Jun 2001, Consulting Engineer, Babcock Borsig Power
Links
Practical information
- Expert
- Free