Digital Twins-in-the-Loop: Integrating Measurement, Modeling, and Cyber-Physical Components of Power Systems

Abstract: Through this talk we dive into a Digital Twin (DT) framework that leverages reduced-order models enhanced with data-driven (aka hybrid fidelity DT) techniques for real-time monitoring and control of cyber-physical devices in modern power grids. It discusses how we can create high-fidelity DTs as cyber-physical replicas by merging dynamic modeling with data-driven estimation, ultimately boosting situational awareness and grid stability. The talk also covers the pros and cons of traditional model-driven, data-driven DT, and hybrid fidelity DT approaches. We will take a look at the software stack and integration pipeline necessary for deploying hybrid fidelity DTs in cloud or fog environments. Key challenges in developing DTs that incorporate reduced-order models and observer-based control schemes to capture system interactions and enhance accuracy will be discussed. Additionally, we will explore data synchronization issues for Digital Twins-in-the-Loop (DTiL), along with real-time data imputation and decentralized control methods, demonstrating how we can tackle measurement gaps and network disturbances. Finally, we will showcase results from DT implementations for heat pumps, EV chargers, and solar inverters, highlighting the emerging role of DTiL in reinforcement learning, federated learning, and spatio-temporal frameworks in the future of electrical <a href="http://grids.

Co-sponsored” target=”_blank” title=”grids.

Co-sponsored”>grids.

Co-sponsored by: School of Electrical Engineering and Computer Science, University of Ottawa

Speaker(s): Javad Fattahi

Room: STE-5084, Bldg: SITE – the southernmost building of the University of Ottawa Main Campus, School of Electrical Engineering and Computer Science, 800 King Edward Ave, Ottawa, Ontario, Canada, K1N 6N5, Virtual: https://events.vtools.ieee.org/m/519732