A Novel State Estimation Method for Smart Grid Under Consecutive Denial of Service Attacks
Xue Li, Cheng Jiang, Dajun Du, Wenting Li, Minrui Fei, Lei Wu
Abstract
Persistent data packet losses induced by consecutive denial-of-service (DoS) attacks could fail traditional state estimation (SE) algorithms that highly rely on the completeness of dataset. To solve the problem, this article explores a novel SE algorithm with enhanced SE accuracy for power systems against consecutive DoS attacks. First, according to the characteristics of data packet losses induced by DoS attacks, we design a strategy by using the latest received measurement packet to compensate for consecutive data packet losses, and reconstruct the power system model. Second, by integrating Holt’s two-parameter exponential smoothing and extended Kalman filter techniques, a new enhanced SE algorithm is proposed, where the statistical properties of data packet losses are contained in the recursion formulas of the state prediction and state filtering processes. Third, the boundedness of estimation error covariance matrix and prediction error are proved. Finally, the proposed algorithm is compared with traditional SE algorithms via three IEEE testing systems and verified in a real power system. Simulation results illustrate the effectiveness and efficiency of the proposed algorithm under various data packet losses scenarios.