Automated Red-teaming for Securing Critical Infrastructures against Instability Attacks

Here we present a novel methodology for end-to-end analysis of instability in power systems. Our methodology uses a modeling representation known as the Dynamical Structure Function (DSF) to identify how linear time-varying attacks propagate through the exposed portions of the system. We show how the DSF of a cyber-physical system can be used as a way to model the attack surface. We present a metric for identifying intrinsic vulnerabilities, and validate our methodology by performing a vulnerability analysis of the IEEE 14-Bus System, including a simulation of an instability attack.