Research And Development Scientist
CurrentResearch and development efforts towards the use of artificial intelligence (incl. statistical machine learning and reinforcement learning), control theory and algorithms, and advanced analytics, in estimating the state of critical energy infrastructure, understanding its health, and taking informed decisions related to safe and efficient operations. Highlights of recent efforts:* Operations- and Maintenance-oriented long term optimization and supervisory control of advanced nuclear reactors, using a reinforcement learning-based approach, appropriately tailored to the nuclear domain peculiarities and to the use of multiscale digital twins for fast, efficient, and safe simulation-based learning (supported by Advanced Research Projects Agency - Energy / ARPA-E) [https://doi.org/10.1016/j.engappai.2022.105454]* Probabilistic assessment of the state of electricity grids heavy on distributed energy resources in the presence of critical loads and distributed / multimodal, using probabilistic graphical models and Bayesian inference techniques for structure and parameter learning (supported by Energy Efficiency and Renewable Energy / Solar Energy Technology Office of U.S. DOE) [https://doi.org/10.1109/ISGT51731.2023.10066364]* Estimation of distributed renewable power generation and uncertainty quantification by fusing real-time environment measurements with physics models and opportunistic signatures using Bayesian neural networks (supported by Office of Electricity of U.S. DOE) [https://doi.org/10.1109/ISGT51731.2023.10066417]* AI/ML-based true load mapping for low-visibility, distributed-renewable inclusive grids (supported by Office of Electricity of U.S. DOE) [https://doi.org/10.1016/j.apenergy.2024.123291]
