A Probabilistic Graphical Model for Predicting Cascade Failures of Electric Vehicle Charging Networks Caused by Hurricanes

Rapid proliferation of electric vehicles (EVs) accelerates the construction of charging stations, evolving into EV charging networks (EVCNs). Extreme weather may trigger the cascading failures in the EVCNs, affecting the stable operation of EV charging stations. This paper proposes an approach for predicting the cascading failures caused by hurricane considering the load shedding and the following EV load redistribution. A Bayesian network based probabilistic graphical model is established to represent the interaction between the outages of charging stations. An AC-based Cascading Failure model (ACCF) is utilized to identify the overload branch and conduct load shedding. The load redistribution is modeled by the effective resistance between the nearby charging stations. In the case study, we predict the probability of failure of EV charging systems to verify the effectiveness of the method by comparing real data for different hurricane scenarios. The result indicates that the load redistribution should not be ignored when predicting the cascading failures of EVCNs.