Plug and Play Detector Design for DC Microgrids With Unknown-Inputs-Based FDI Attack

Abstract

DC microgrids, due to their deep integration of control, computing, communication technologies, and physical equipment, are susceptible to cyber-attacks. Consequently, this paper is dedicated to the development of a novel attack-defense framework for generalized DC microgrids. Firstly, an unknown-inputs-based false data injection (FDI) attack strategy is studied from the adversary’s perspective, unlike traditional stealthy attacks requiring non-minimum phase zeros or unstable poles, which conceals the attack signal as false unknown inputs (FUI) to maliciously disrupt current sharing and voltage balancing. Secondly, a comprehensive analysis of the stealthiness and destructiveness of FUI attack is provided, and a dual-observer-based detector is well constructed to detect the FUI attack and isolate the compromised distributed generation units. Then, structured Lyapunov matrix and semidefinite programming are ingeniously employed to solve the distributed observer gains simultaneously. Moreover, plug and play (PnP) performance is also analyzed to ensure the scalability of proposed FUI attack detector. Finally, the destructiveness and stealthiness of proposed FUI attack, as well as the effectiveness of designed detection scheme are demonstrated through simulations using MATLAB/SimPowerSystems Toolbox.