IDAT: An Impact-Driven Adaptive Threshold for Detecting False Data Injection Attacks on Wind Turbines

The increasing deployment of wind power systems has raised concerns about their vulnerability to cyberattacks due to remote locations and insecure communication protocols. False data injection (FDI) attacks can bypass detection mechanisms and manipulate system parameters, which may lead to inefficient operation or even failure of wind turbines. In the face of FDI attacks, the first step is to achieve attack detection. However, the strong nonlinearity of wind power systems leads to a difficult problem in detection threshold selection. To this end, this paper proposes a novel impact-driven adaptive threshold (IDAT) selection methodology to detect FDI attacks on variable-speed wind turbines. Firstly, quantitative and qualitative conclusions are obtained based on in-depth analysis of attack impacts. Then, the framework of a general observer-based detector is analyzed, and the difficulties that may arise during threshold selection are explained. Finally, the completeness condition of the detection mechanism is established, and the threshold is adaptively determined in accordance with the specific impact of each attack, ensuring high interpretability. Extensive simulations are conducted on a 5 MW doubly-fed induction generator-based wind turbine to demonstrate the effectiveness of the proposed approach.