Blending Data and Physics Against False Data Injection Attack: An Event-Triggered Moving Target Defence Approach

Fast and accurate detection of cyberattacks is a key element for cyber-resilient power system. Recently, data-driven detectors physics-based Moving Target Defences (MTD) have been proposed to detect false data injection (FDI) attacks on state estimation. However, the uncontrollable positive rate detector extra cost frequent MTD usage limit their wide applications. Few works explored overlap between these two areas. To fill this gap, paper proposes blending approaches enhance performance. start, physics-informed attack identification algorithm proposed. Then, an protocol triggered by alarm from detector. The formulated as bilevel optimisation robustly guarantee its effectiveness against worst-case around identified vector. Meanwhile, hiddenness also improved so that defence cannot be detected attacker. feasibility convergence, convex two-stage reformulation derived through duality linear matrix inequality. simulation results verify physics can achieve extremely high while simultaneously reducing MTD. All codes are available at https://github.com/xuwkk/DDET-MTD .