Neural Network-Based Resilient Control of a Wind Turbine Subject to False Data Injection Attack

The importance of sustainable energy has grown significantly in recent years, with wind energy receiving particular attention. However, wind turbines are increasingly vulnerable to cyber threats, especially those targeting control protocols and cyber-physical subsystems. In response, resilient control has emerged as a promising approach to enhance system security. Nevertheless, implementing such strategies in wind turbines poses challenges, particularly in maintaining finite-time stability and ensuring consistent power output. To address these issues, this paper proposes a novel control framework capable of detecting and mitigating anomalies, thereby ensuring the safe and stable operation of wind turbines. Specifically, a resilient pitch angle control method is developed using a combination of a Radial Basis Function Neural Network (RBFNN) and an adaptive sliding mode controller. The main objective is to regulate the pitch angle to maintain the rated output power while countering cyber threats. Stability is rigorously analyzed through Lyapunov theory, ensuring finite-time convergence of the tracking error. Numerical simulations based on practical wind turbine data confirm the effectiveness of the proposed control structure in compensating for both false data injection (FDI) attacks and external disturbances.