Data-Driven False Data Injection Attacks on State Estimation in Smart Grid

In this paper, we focus on false data injection attacks (FDIAs) on state estimation in smart grid. Without topology information and parameters of smart grid, two kinds of Data-Driven FDIAs (DDFDIAs) are constructed with noisy measurements, which can keep stealth in the residue-based bad data detection (BBD) of state estimator. Furthermore, the proposed DDAs are more lower cost and sparser due to the limited energy of the attackers. Finally, the simulation experiments are achieved on 30-bus power system. Results show that the constructed DDFDIAs are feasible and undetected by BBD.