Detection and Defense Against Multi-Point False Data Injection Attacks of Load Frequency Control in Smart Grid
Xing‐Chen Shangguan, Ming-Hui Yu, Chuan‐Ke Zhang, Yong He
Abstract
Load frequency control (LFC) systems operate over open communication networks, making them susceptible to cyber threats such as false data injection (FDI) attacks. These attacks, especially when launched from multiple points, present serious risks to the stability and security of the LFC system by corrupting critical measurement data. This paper proposes novel multipoint FDI attack detection and defense methods based on temporal analysis neural networks to increase the resilience of LFC systems under attack. Specifically, a data-driven dual-layer long short-term memory (DL-LSTM) attack detector is designed for efficient detection of multipoint FDI attacks. The method leverages a hierarchical LSTM architecture to capture temporal dependencies effectively and identify attack patterns within time series data. Moreover, a bidirectional gated recurrent unit (BiGRU)-based defense method is proposed to mitigate the impact of such attacks, leveraging its superior performance in time series data prediction to ensure effective postattack defense. The proposed detection and defense mechanism is validated through simulations on a smart grid with three-area LFC systems, incorporating the practical complexities of high-voltage direct current (HVDC) links, renewable energy sources (RESs), and system nonlinearities, demonstrating its effectiveness in attack detection and its promising applicability to complex power grid environments.