A stochastic model for cascading failures in smart grid under cyber attack

The coupling of power systems with cyber networks poses security issues that may threaten the safety and robustness of power systems. In this paper, by combining stochastic process with a state transition description, a stochastic model is presented for describing cascading failure in a cyber-coupled smart grid. A dedicated algorithm is developed to simulate the dynamic profile of the cascading failures, with consideration of the effect of power overloading, malware contagion and interdependency between the power grid and cyber network. Stimulation results on two test cases show that the dynamic profile of the cascading failures in the coupled system displays a 'staircase-like' pattern, which can be interpreted as a combined feature of the typical step propagation profile triggered repeatedly by cyber attacks due to network coupling. Furthermore, compared with the isolated power system without coupling cyber network, both the extent and rapidity of power blackouts were intensified by the cyber coupling especially under the cyber attack.