Blending Data and Physics Against False Data Injection Attack: An Event-Triggered Moving Target Defence Approach
Lightning Talk
cyber-physical power system; FDI attacks; attack detection and identification; moving target defence; event triggering
Recently, data-driven detectors and physics-based Moving Target Defences (MTD) have been proposed to detect false data injection (FDI) attacks on power system state estimation. However, the uncontrollable false positive rate of the data-driven detector and the extra cost of frequent MTD usage limit their wide applications. Few works have explored the overlap between these two areas to collaboratively detect FDI attacks. To fill the gap, this paper proposes blending data-driven and physics-based approaches to enhance the detection of FDI attacks. To start, a physics-informed data-driven attack detection and identification algorithm is proposed. Following this, an MTD protocol is triggered by a positive signal from the data-driven detector and is formulated as a bilevel optimisation to robustly guarantee the effectiveness of MTD against the worst-case attack around the identified attack vector while maximising the hiddenness of MTD. To guarantee the feasibility and convergence, the bilevel nonconvex optimisation is separated into two stages, and for each stage, a semidefinite programming is derived through duality and linear matrix inequality. The simulation results verify that by blending data and physics, it can significantly enhance the detection performance while simultaneously reducing the false positive rate of the data-driven detector and the usage of MTD.