Detection of Stealthy Cyber Intrusion in Smart Electric Grid Using Advanced State Estimation

Security of the electric grid is crucial for the efficient, smooth, and reliable functioning of a power system. The most common points where the attacker can compromise the security are the communication infrastructure and meters. Once these are compromised the attacker can inject data evading Bad Data Detection (BDD) algorithms and this type of stealthy attack is known as Undetected False Data Injection (UFDI) attack. The information regarding power grids is highly secretive and these pose a limitation in terms of knowledge of line, information on secure points for attackers. The model used in this paper can be used to detect stealthy attacks which are modeled using constraints such as knowledge of line, availability, accessibility, and security to yield a satisfiable attack vector. The proposed model is demonstrated and evaluated on IEEE 5-bus and 14-bus systems and it shows the attack vectors satisfying the input constraints given. The same can be used for various other large bus systems as well. Z3 SMT solver is used for solving the satisfiability constraints.