Fast Screening of Vulnerable Transmission Lines in Power Grids: A PageRank-Based Approach

In power system operation, there are usually a few vulnerable lines (VLs) which are prone to exacerbate the propagation of cascading failures and lead to severe blackouts with high probability. To facilitate preventing the power grid from cascading blackouts, in this paper, a PageRank-based fast screening method is derived to identify VLs. To this end, we extend the concept of interaction graph to model cascading outages as a directed weighted graph, where impacts of hidden failures are first taken into account. Based on the graph, we modify the celebrated PageRank algorithm to derive an ultra-fast VLs screening algorithm for large-scale power grids. Simulations carried out on the IEEE 118-bus system show that the proposed method can achieve high-accuracy VLs screening while the speed can be enhanced for several orders of magnitude compared with cascading failure simulation methods. Tests on a 1122-bus real power grid in China demonstrate that it can discriminate the most VLs in a practical power grid within only several minutes, showing its great potential for on-line application.