Quickest Detection of False Data Injection Attacks in Smart Grid with Dynamic Models

Abstract

A quickest intrusion detection algorithm is proposed to detect false data injection attacks (FDIAs) in smart grids with time-varying dynamic models. The quickest detection algorithm aims at minimizing the worst case detection delays (WDDs) of cyberattacks, subject to an upper bound of the false alarm rate. Since power-grid state transitions could be caused by either cyberattacks or sudden change in loads or grid configurations, we propose to distinguish between an FDIA and a sudden system change by using a time-varying dynamic model, which can accurately capture the dynamic state transitions due to changes in system configurations. A dynamic state estimation algorithm is developed to estimate and track the time-varying and nonstationary power-grid states. The quickest detection algorithm is developed by analyzing the statistical properties of dynamic state estimations, such that the algorithm minimizes the WDD while accurately distinguishing FDIA from sudden system changes. A Markov-chain-based analytical model is used to identify the detector’s parameter and quantify its performance. Simulation results demonstrate that the proposed algorithm can accurately detect and remove false data injections or system faults with minimum delays. The proposed algorithm can be implemented to harden intelligent electronic devices (IEDs) or supervisory control and data acquisition (SCADA) systems to improve their resilience to cyberattacks or system faults, thus improving the cybersecurity of smart grids.