Cyber Attack Detection and Isolation for Smart Grids via Unknown Input Observer

This paper studies the detection and isolation problem against False Data Injection Attack (FDIA) for distributed smart grids, by using the designed Unknown Input Observer (UIO). Firstly, a simple but classical mathematics model of distributed smart grid is introduced. Considering the covert of FDIA, the Unknown Input Observer is designed for detecting and isolating cyber attack in smart grids. Then, considering the disturbance and uncertainty of grid system model, the adaptive thresholds are computed for detecting the cyber attack. Furthermore, UIO-based attack detection and isolation Algorithm is proposed, by computing the observed residuals. Finally, simulations are given to illustrate the effectiveness of the proposed algorithm.