Non-linear state recovery in power system under bad data and cyber attacks

• DOI: 10.1007/s40565-019-0561-2
• Publication Date: 2019-07-30T07:02:47Z
• Authors: Ali TAJER; Saurabh SIHAG; Khawla ALNAJJAR
• Keywords: Isolation; Gaussian Noise

The problems of recovering the state power systems and detecting instances bad data have been widely studied in literature. Nevertheless, these two operations designed optimized for most part isolation. Specifically, estimators are based on minimum mean-square error criteria, which is only optimal when source distortions Gaussian random noise. Hence, fail to perform optimality further contaminated by data, cannot necessarily be modeled additive terms. problem estimation has extensively. But fundamental performance limits attendant decision rules unknown potentially compromised (due sensor failures) or structured cyber attacks, also referred false injection attacks). This paper provides a general framework that formalizes underlying connection between detection routines. We aim carry out combined tasks presence form accurate estimations grid. characterizes detectors estimators. Furthermore, gains with respect existing established through numerical evaluations.