Simulation of dynamical systems is a principal tool for modeling and studying a wide range of complex physical phenomena of scientific interest or industrial value. Examples include structural dynamics, biological systems, fluid dynamics, and power and gas networks. The need for accuracy in the modeling stage leads to large-scale dynamical systems and simulating them presents unmanageably large demands on computational resources. Alleviating this computational burden is the main motivation for model reduction.
Professor Gugercin's research focuses on developing theoretical analysis and computational framework for constructing high-fidelity reduced models, that are much easier to simulate yet approximate the original system accurately for a wide range of operating conditions. In some cases, these reduced models are constructed purely from data as in the setting of data-driven modeling.
This research area is highly interdisciplinary in character and covers topics such as numerical analysis, systems and control theory, numerical linear algebra, approximation theory, and optimization.