Mitigating Concurrent False Data Injection Attacks in Cooperative DC Microgrids

Abstract

Limited global information in dc microgrids with distributed cooperative control makes them vulnerable to cyber attacks, which can lead to their destabilization and shut down. Here, we discuss a novel false data injection attack (FDIA) model, termed as the concurrent attack, that can compromise local and communicated estimated voltages simultaneously. We formalize that such an attack could be disguised as a conventional FDIA on the estimated voltages transmitted in communication links (termed as the communication link attack), thereby masking the presence of the attack on local estimated voltages and rendering corresponding mitigation attempts ineffective. Second, we present an energy-based detection strategy based on the intrinsic mode functions obtained using the ensemble empirical mode decomposition method. Further, a differentiation criterion using the voltage correction terms generated from the voltage observer is employed to help distinguish between the concurrent attack and the communication link attack. An event-driven mitigation strategy is then used to replace the attacked signal with a reconstructed signal. Finally, the efficacy of the proposed resilient control scheme is demonstrated using both simulations and experimental results.