Convex Optimization of Cyberattacks Overflowing Multiple Lines in Cyber-Physical Power Systems

Abstract

Cyberattacks designed to overflow transmission or distribution lines may lead to cascading failures or blackouts in smart grids. Accurately modeling these attacks is the first step in designing appropriate countermeasures. Therefore, this article aims at power system vulnerability analysis from the operator’s perspective, i.e., identifying the scenarios, where false data injection (FDI) cyberattacks can lead to overflow of multiple lines. A nonlinear optimization model is developed for cyberattacks that can bypass the state-estimation and result in congestion of multiple lines in smart grids. Unlike existing line congestion attacks that focused on dc state-estimation, the proposed model is based on FDIs on targeted set of buses to overflow targeted set of lines by bypassing ac state-estimation. Considering ac state-estimation in the attack model generalizes the proposed model for both transmission and distribution-level power grids. The nonconvex model is then converted to a convex formulation using second-order cone programming relaxation, and global optimality conditions are discussed. Several case studies are considered to validate the effectiveness of the proposed attack model using IEEE 118-bus benchmark.