Resilient AC Microgrids Against Correlated Attacks
Shan Zuo, Deepak Pullaguram, Mohamadamin Rajabinezhad, Frank L. Lewis, Ali Davoudi
Abstract
Multi-inverter AC microgrids increasingly rely on local embedded controllers and distributed communication networks to meet operational requirements, which makes the microgrids vulnerable to physical and cyber attacks. Conventional resilient control strategies generally assume that the attack signals are bounded and uncorrelated. In this paper, we study the ramifications of allowing the antagonistic inputs to be unbounded and correlated. We consider a two-layer hierarchy for networked multi-agent systems with two opposing teams on different directed communication graphs: a control protagonist team with cooperative multi-inverter microgrids and an attack antagonist team with interacting attackers. We consider three types of unbounded attack injections launched from the antagonist layer, namely, coordinated and correlated attacks on the sensor measurements, as well as generally unbounded attacks on actuator commands and communication channels. We propose a fully distributed control framework to guarantee uniform ultimate boundedness for the secondary frequency regulation and voltage containment of AC microgrids against malicious attacks. The proposed results are validated on a modified IEEE 34-bus test feeder system, which is emulated in a controller/hardware-in-the-loop environment.