State Estimation in Power System under Deterministic False Data Injection Attack Using Minimization of Nuclear Norm and ℓ<inf>1</inf>Norm with Noiseless Constraint Substitution

Abstract

The use of digital technology causes the susceptibility to a malicious data attack. A careful mitigation of such unobservable false data injection (FDI) is vital for monitoring the electric power distribution health. This work considers a state estimation problem under a FDI stealth in modern power systems, e.g., a smart grid network. We propose an improved optimization technique for extracting the attack-free component from the attacked data derived from the measurement and a concept of estimating the power system state vector from the extracted signal component. Numerical simulation is conducted to demonstrate the possible operation of the proposed notion. Numerical results indicate that our proposed technique for extracting the attack-free signal component consumes less computational time than the former work.