Our group specializes in computational mathematics (e.g., optimization, control, statistics, graph theory).
We develop theory, algorithms, and software to tackle problems that arise in diverse scientific and industrial applications (e.g., energy, sustainability, materials).
Scalable Modeling and Optimization
Creating, analyzing, and solving optimization models are core activities in systems engineering.
We are developing graph-based modeling abstractions as well as scalable algorithms and software that facilitate the analysis and solution of complex optimization problems.
We are targeting diverse problems that arise in energy infrastructures, supply chains, and chemical processes.
Data Science and Machine Learning
We are developing algorithms and software for analyzing complex datasets that arise in materials, catalysis, manufacturing, and infrastructures as well as for developing sensing and control technologies.
Here, we are combining techniques from space-time statistics, topology/geometry, graph theory, linear algebra, and machine learning to extract feature information that is encoded in complex data sets.
We are also developing Bayesian optimization techniques to help guide experiments and data collection.
We are developing optimization models and algorithms to tackle challenges that arise in the management of agricultural and plastic waste.
Our models aim to resolve conflicts between economic and environmental goals arising in these systems. Moreover, our models capture complex interdependencies between industrial sectors as well as to provide insight into policy and market incentives that can help facilitate the deployment of more sustainable technologies and management strategies.
We are analyzing economic and resiliency benefits for the power grid that result from the provision of flexibility by established and emerging technologies such as batteries, manufacturing facilities, buildings, and data centers. This flexibility is key to manage large amounts of intermittent and non-dispatchable renewable power and can help withstand severe contingencies that might arise from climate change and cyber attacks.
We are developing optimization and control formulations that capture revenue streams for technologies at multiple scales from day-ahead, real-time, and frequency regulation markets and that seek to determine optimal technology locations that maximize revenue. We perform high-fidelity simulations using physical models and real power grid data. We are also developing optimization formulations to quantify gains in power grid flexibility that arise from widespread deployment of technologies.