Measurement protection to prevent cyber–physical attacks against power system State Estimation

Abstract

Smart applications supporting modern power systems are susceptible to cyber–physical attacks, particularly False Data Injection attacks that manipulate the input measurements of State Estimation (SE) compromising its output states. This paper proposes an Integer Linear Programming formulation that protects an optimal number of measurement units to prevent cyber–physical attacks, enhancing the robustness of SE. Our approach exhibits low complexity, applies to both linear and nonlinear SE, and converges rapidly toward the optimal solution. The formulation requires information about the grid topology and measurement distribution but does not depend on the power flow equations. Also, the generalized formulation can be customized to consider distinct protection costs for all measurement types, various priorities for different measurements, and a range of measurements and pseudo-measurements. Simulations are performed on the widely used IEEE 14 and 118-bus systems to verify the approach for linear and nonlinear SE and illustrate its practicality.