A Bayesian Network Model for Predicting Outages of Distribution System Caused by Hurricanes

To enhance the grid resilience against hurricanes, the anti-disaster preparations should be carried out. As a leading action, an accurate outage prediction under hurricanes is highly important. In this work, an outage prediction method based on the Bayesian network (BN) is proposed. Since the outage is resulted from topology failures due to the hurricane impacts, both the grid topology and the hurricane dynamics are treated as causalities in the proposed BN. The historical data are used to emulate outage samples for training the model. Then, the BN can be used in predicting the outage range with a hurricane forecast. The modified IEEE 14-bus system placed in a real city is tested with two real hurricane events. Test results validate the effectiveness of the proposed method and provide further insights into the grid resilience.