Resilient Control of DC Microgrids Against Cyber Attacks: A Functional Observer Based Approach

Abstract

Direct current microgrids (DCMGs) are swiftly moving toward the realm of communication-dependent distributed cooperative control strategies. The incorporation of cyber layer for robustness, scalability, and reliability makes the system vulnerable toward cyber-attacks. The extent of damage caused by these attacks on DCMG is substantial, to the point where ceasing the operation may become necessary. This article proposes a resilient strategy for the detection and mitigation of the most prominent false data injection attacks (FDIAs) on actuators of nodes of DCMG. An accurate error-free detection is guaranteed using a state machine-based model that makes use of a dynamic signature function that monitors the actuator signal and its estimated value. Linear functional observer (LFO) based mitigation scheme is proposed, in which the affected node is switched to LFO upon a true attack detection. The proposed technique is consistent during transients. Experimentation and simulation studies are carried out for various practical situations for a four-node DCMG to validate the proposed theory.