PROJECT TITLE :

Attack Vulnerability of Power Systems Under an Equal Load Redistribution Model - 2018

ABSTRACT:

This Project studies the vulnerability of flow networks against adversarial attacks. In specific, take into account a power system (or, any system carrying a physical flow) consisting of N transmission lines with initial loads L one, . . . , L N and capacities C 1 , . . . , C N , respectively; the capability C defines the most flow allowed on line i. Beneath an equal load redistribution model, where load of failed lines is redistributed equally among all remaining lines, we tend to study the optimization problem of finding the simplest k lines to attack so as to reduce the number of alive lines at the steady-state (i.e., when cascades stop). This is completed to reveal the worst-case attack vulnerability of the system still on reveal its most vulnerable lines. We have a tendency to derive optimal attack ways in several special cases of load-capability distributions that are practically relevant. We tend to then consider a changed optimization problem where the adversary is additionally constrained by the whole load (additionally to the number) of the initial attack set, and prove that this downside is NP-laborious. Finally, we have a tendency to develop heuristic algorithms for selecting the attack set for each the initial and changed problems. Through intensive simulations, we tend to show that these heuristics outperform benchmark algorithms underneath a wide selection of settings.