Fast, Online, and Efficient Computational Tools for Smart and Resilient Infrastructure Systems
Speaker: Negin Alemazkoor, University of Virginia
Abstract: Optimal management of critical infrastructure systems directly affects the welfare of society. Achieving optimal management requires fast, reliable, and efficient analysis of the system. However, such analysis is extremely challenging due to the large size of infrastructure systems and various inherent sources of uncertainty. This talk focuses on how advances in computing and sensing technologies can facilitate the analysis and management of large infrastructure systems. In particular, I present some of the computational tools that I have recently developed that a) accelerate computations for probabilistic system analysis and uncertainty quantification; b) accurately capture the changes in demand trend and/or system topology in an on-line fashion and update the prediction models accordingly; and c) efficiently collect system data to alleviate the cost of data storage and communication. I also present some applications of these computational tools in critical infrastructure systems, such as transportation and power systems, and discuss the computational cost savings and accuracy improvements.
Bio: The overarching objective of Negin Alemazkoor's research is to develop sensing and computing methodologies for fast and reliable analysis of smart and interconnected infrastructure systems under uncertainty. Her research is inherently multi-disciplinary and aims to advance the reliability and resilience of infrastructure systems by facilitating fast and accurate system analysis that leads to optimal system operation and management.