Dynamic State Estimation of Power System Under FDI Attacks Based on Interpolating AHEKF Algorithm

Abstract

With the application of communication network technology in power systems, traditional power grids have gradually developed into cyber-physical systems (CPS), and false data injection (FDI) attacks have become a hidden danger that affects the operation of power CPS. To solve the problems of low estimation accuracy and model uncertainty when the extended Kalman filter (EKF) algorithm is subjected to FDI attacks in power systems, an interpolation adaptive $\mathrm{H}_{\infty}$ extended Kalman filter (IAHEKF) algorithm is proposed for the dynamic state estimation of power systems under FDI attacks. The new algorithm uses an interpolation strategy to reduce the linearization error of the EKF algorithm and introduces adaptive $\mathrm{H}_{\infty}$ theory to update the error covariance, minimizing the error upper bound caused by model uncertainty; moreover, it uses the Sage-Husa estimator to calculate the noise covariance, reducing the impact of unknown noise on state estimation. Finally, tests are conducted on the IEEE-14 node system and IEEE-30 node system, and the results show that the IAHEKF algorithm has higher estimation accuracy under different attack scenarios.