Skip to Main Content
Transient stability can be rapidly assessed using the artificial intelligence technology. In this paper, a fast transient stability assessment method based on Bayesian network classifier was proposed from the perspective of data mining. First, select the characteristic quantities which reflect the power system transient process rapidly as the attribute variables of the Bayesian network classifier, then determine the stable event's posterior probability using of the prior information and sample data which is produced massively by numerical simulation algorithm. When the disturbances occur, we can judge the power system is stabile or not by reasoning according to the corresponding attribute variables. Because any classifier has the probability of misclassification, the boosting algorithm of Bayesian network classifier is applied. Finally, we conduct a numerical simulation on New England 39-bus system to verify the effectiveness of the classifier.