Abstract
The aim of this work is to present a joint model for evaluating the impact of targeted attacks to power systems considering cascading and hidden failures using centralities. The targeted attacks are based on importance- based selection of links using power traffic centrality measures on line graphs. The algorithm presented here is based on the estimation of satisfied demand on stages of cascading failure, and the probability of presenting hidden failures is modeled as a Poisson random variable. Simulations were carried out in four test systems, evaluating different number of hidden failures per trial and different centrality measures for selection of target branches. Results indicated that augmenting the number of hidden failures increases the possible unsatisfied demand after the cascading failure, which was not considered in a traditional approach. The selection of target branches with degree centrality was consistent an allowed to evaluate the largest impact on the test systems. The model can be applied to assess the impact of cascading and hidden failures employing parameters based on failure records of the power system. This approach can help to deeply assess the vulnerability of the power system when subject of intentional attacks to the most important elements based on centralities.