Grant: Alliance Grants

The proliferation of diverse power devices such as renewable generation systems and the presence of ubiquitous communication in the modern power grid introduce many advantages along with new vulnerabilities and grid reliability issues. Perturbations stemming from natural events (e.g. faults, environmental conditions, power quality, etc.) and adversarial cyber-attacks (e.g. false data injection attacks) must be readily detected and addressed to preserve the stable operations of the electric grid. The proposed research program aims to design, implement and evaluate a suite of intrusion detection systems (IDSs) that aim identify and report grid perturbations in a timely manner. Specifically, we focus on data-driven methods for developing these IDSs. For this, a comprehensive set of data points will be curated using real-time digital simulators and generative techniques which are then utilized to train the proposed IDSs. Additional considerations, such as the ability of the IDS systems to adapt to changes in the grid infrastructure and preserving privacy when utilizing grid measurements for training and inferencing purposes, will be incorporated to develop novel adaptive, distributed and privacy-preserving IDSs. This research program supports national security by ensuring the reliable operations of the power grid along with climate change initiatives such as Clean Energy Regulations developed by the federal government in which grid reliability is a key element.