XGBoost based fake data injection attack detection method for power grid

Different from distributed denial of service attack, fake data injection attack (FDIA) can bypass the traditional residual-based bad data detection, and affect the accuracy of state estimation by tampering with the power grid information collected by SCADA, thus posing a threat to the security of power grid. Aiming at the state estimation of the smart power grid, an attack detection method based on XG-Boost is proposed. By training and adjusting the parameters of the XG-Boost model, the characteristics of the data are extracted, and a high accuracy false data injection attack detection model is realized. Finally, the simulation is carried out with the open-source data set from Mississippi State University, and the experimental results show that the proposed method is accurate and effective.