Simulation and Automated Anomaly Detection for Power Grids

There is growing concern over the potential for a cyberattack on the electric power grid’s operational technology (OT) systems, an event that could break the backbone of all America’s economic sectors. Software to provide situational awareness to a grid operator during a critical cyberattack, including during black start restoration attempts, requires three components: a robust and reliable state estimation that uses available measurements to generate a model of the system, anomaly detection that flags compromised measurements with untrustworthy grid data, and a robust power flow engine to simulate and adjust the crank path on the fly. The proposed project is to build commercial-grade software to achieve each of these functionalities leveraging advanced simulation and modeling approaches adapted from the electronic design automation industry. The target product will have applications for both cyberattack prevention and recovery, and its physics-based foundation naturally complements existing network security systems to provide defense-in-depth against grid-focused cyberattacks. The first twelve months of the project will involve development, testing, and benchmarking against real utility data, followed by a second twelve-month period to pursue real-time integration with a partner utility.