A Graph Theory-Based Approach to Detect False Data Injection Attacks in Power System AC State Estimation

Nowadays, an energy management system (EMS) located at power grid control centers plays a critical role in ensuring the reliable operation of a grid. The SE serves as the main interface of EMS with the power grid, and because of that other EMS functionalities rely on its provided data. Recent studies show that the state estimator is vulnerable to various types of cyber-attacks. Among different types of such attacks, the false data injection attacks (FDIAs) are proven to be one of the most stealth ones and attracted much attention. Various countermeasures have been proposed to detect such kind of attacks. However, these approaches are generally based on some assumptions, such as being limited to detecting a specific type of FDIAs or the use of dc formulation for modeling the attack. In this article, a new algorithm is proposed to detect the FDIAs on ac state estimation. This algorithm applies outlier detection techniques on the state estimation results and then detects the possible attack by a graph-theory-based approach. The performance of the proposed algorithm is evaluated by simulating two main types of FDIAs in the IEEE 14 bus and 118 bus test systems. The results show that the algorithm performs satisfactorily in detecting both kinds of FDIAs with high rates.