Kalman Filtering Based Interval State Estimation For Attack Detection

Abstract The accurately estimated energy state is not only critical for the normal operation of power system, but also is the foundation of power system dispatch, control and assessment. To maintain the accuracy of the estimated state, bad data detection (BDD) is utilized by power system to get rid of erroneous measurements based on meter failures or outside malicious attacks. However, many studies have revealed that hackers can circumvent BDD and inserted false data into the power systems estimated state. For the sake of precise and effective cyber-attack detection, an original detection mechanism based on Extended Kalman filter interval state estimation (ISE) is proposed in this paper. In the detection mechanism, whose objective is to assess its maximum and minimum boundary values around the state variables. Finally, the effectiveness of Kalman filter interval state estimation method is presented in nonlinear modes to verify on standard IEEE 14-bus system.