Robust Bayesian Regression Model of Centrality and Voltage Stability Index for Power Networks under Nodal Attack

Abstract

Electrical node centrality for the power networks is an essential parameter to identify the critical nodes under attack. Topological analysis is vital for evaluating the network robustness while electrical characteristics have to be considered to make the analysis consistent for realistic power networks. However, the capacity limit of the power network changes under various nodal attacks. It is essential to find the relationship between the loading margin limit of the power network with the node centrality features, so that appropriate measures can be considered to improve the robustness of the power networks. Thus, voltage stability index (VSI) is defined for every node, and its centrality features are modelled. Robust Bayesian regression is used to model the nodes responsible for a change in loading margin and causing grid blackout. The method has been validated on benchmark complex power networks like reduced Great Britain network, IEEE 57-bus and IEEE 118-bus systems.