부산대학교 도서관

The strong interplay between the power grid and the corresponding communication and control network plays a pivotal role in the resilience of the smart grid. In this paper, the dynamics of the interdependence among smart-grid subsystems such as the power grid, communication network, and response of human operators are captured during the propagation of cascading failures. A previously developed Markov-chain based model is refined into an interdependent Markov chain model to capture the role of cyber threat from the communication network and the human-operator error during cascading failures. The state transitions of the Markov chain are parameterized by the critical operating parameters of the power grid. The calculations assume a generic form of correlation between the level of and damage from cyber-attacks, on the one hand, and the level of interdependence on the other hand. The model finds the optimal level of interdependence, i.e., the trade-off between well-informed control and vulnerability to attacks that minimizes the probability of massive cascading failures in power grids. There is a point of diminishing returns beyond which the harm of exposure to cyber threat outweighs the benefits of information.