Title: TrustML in Power Grids: Security, Robustness, and Statistical Risks

Abstract: AI/ML advances are increasingly transforming the paradigm in the operations and control of the electric power grids. Nonetheless, as a critical infrastructure, power grids need to ensure the AI/ML solutions are as trustworthy as they are intelligent. In this talk, I will discuss a few recent research directions to achieve reliable, secure, and risk-aware applications of AI/ML to power grids. Our ICML work, LEVIS: Large Exact Verifiable Input Spaces for Neural Networks, will be highlighted. LEVIS aims to analyze the robustness of neural by identifying the largest reliable input space. This can greatly extend the capability of existing robustness verification methods that only assess the worst-case output for a given input space. We introduce two LEVIS approaches and develop a scalable solution via a complementarity-constrained (CC) reformulation, achieving up to a 6 times runtime reduction. We validate the applications of LEVIS to optimal power flow decision prediction and image classification, with demonstrated performance gains and geometric insights into the verifiable region.

Bio: Hao Zhu is an Associate Professor of Electrical and Computer Engineering (ECE) at The University of Texas at Austin. She received the B.S. degree from Tsinghua University in 2006, and the M.Sc. and Ph.D. degrees from the University of Minnesota in 2009 and 2012, all in electrical engineering. From 2012 to 2017, she was a Postdoctoral Research Associate and then an Assistant Professor of ECE at the University of Illinois at Urbana-Champaign. Her research focus is on developing algorithmic solutions for learning and optimization problems in future energy systems. Her current research interests include machine learning for power system operations and resilience enhancements, and the nexus between power grids and the growth of cyber/computing technologies, especially GenAI workloads. She is a recipient of the NSF CAREER Award and the faculty advisor for four Best Student Papers awarded at the North American Power Symposium. She is currently an Editor for the IEEE Transactions on Smart Grid, and IEEE Transactions on Control of Network Systems.