False Data Injection Attack Detection Based on Wavelet Packet Decomposition and Random Forest in Smart Grid
Document Type
Conference Proceeding
Publication Date
1-1-2022
Abstract
As one of the critical infrastructures, the safety and reliability of the smart grid are directly associated with the development and stability of society. However, studies have shown that the power grid is at risk when the parameters are manipulated and cyber-attacks are generated against the state estimation, i.e., under false data injection attack (FDIA). Currently, a rich body of literature has studied on the FDIA defense methods, but most of them focus on the direct current (DC) scenario. This paper proposes a novel detection model that combines the wavelet packet decomposition (WPD) technique with the random forest (RF) algorithm. The WPD is able to capture the deviation of parameters from the normal conditions, whereas the RF is developed to classify these features and effectively identify the malicious data. The proposed model is also evaluated using real-world data on IEEE 118-bus power system. The results demonstrate excellent performance on precision rate and recall rate under varying scenarios.
Recommended Citation
Chen, Zhenyu; Yuan, Shuai; Wu, Longfei; Guan, Zhitao; and Du, Xiaojiang, "False Data Injection Attack Detection Based on Wavelet Packet Decomposition and Random Forest in Smart Grid" (2022). College of Health, Science, and Technology. 819.
https://digitalcommons.uncfsu.edu/college_health_science_technology/819