Identifying core nodes in interaction graphs for critical component analysis in cascading failures of power grids

This paper introduces a novel approach to identifying critical lines in power systems during cascading failures, addressing significant limitations in previous methodologies such as computational inefficiency and limited effectiveness. The proposed methodology is inspired by social network analysis techniques for evaluating the importance of nodes and determining core roles within a network. By adopting these techniques, multiple centrality metrics – degree, ego-betweenness centrality, and eigenvector centrality – are applied to assess the importance and role of lines in interaction graphs of cascading failures. The methodology can be applied to any type of interaction graph of cascading failures and involves selecting centrality metrics with strong correlations to outline the particularity of critical lines. The importance of each line is evaluated as the normalized sum of its three metrics. Critical lines are identified based on their deviation from the statistical correlation of the overall interaction graph, analogous to the role determination process in social networks. The effectiveness of the proposed approach is demonstrated through extensive testing on IEEE 39-bus and 118-bus systems. The identified critical lines are validated with a method from the literature and by analysing the effects of upgrading the critical lines, demonstrating the accuracy and reliability of the proposed methodology. By leveraging interaction graphs and simulation data, our approach provides a robust framework for mitigating cascading failures. The results indicate that this methodology not only improves computational efficiency but also enhances the precision of critical line identification, making it highly suitable for real-time applications in power system stability and reliability.