A Prediction Model Towards the Cascading Failure of Power Grids

Power grids are closely related to people's daily life, and cascading failure plays a vital role in their reliability, which has attracted intensive research interests in the past few years. Numerous researches concerning cascading failure in dynamic power grids, still, most of them focus on the prediction of cascading failure and recovery after failure while Time-of-Arrival (TOA) of cascading failure is rarely concerned. The TOA -based prediction of cascading failure often needs to consider the entire cascading process of the target network during the entire cascading failure process, which brings great loss in network reliability and computational overhead. In light of this, we propose a prediction model of cascading failure based on TOA to implement the prediction of cascading failure in dynamic power grids. We successfully perform our model on two artificial synthetic networks and one real-world network by comparing against four widely used prediction models. We further analyze the non-negligible superiority of our scheme overall benchmarks. As far as we know, we are the first to introduce the conception of TOA for cascading failure prediction in the dynamic power grids.