Modified Bayesian Networks-based Insulation Fault Diagnosis for Power Equipment
In order to ensure safe and efficient operation of power system,achieve accurate diagnosis of insulation faults,and improve the reliability of power equipment,a new insulation fault diagnosis method is proposed based on a modified Bayesian network.This methodology entails the power equipment status signal sampling by sensors,the determination of characteristic parameters and the extraction of fault characteristics by statistical feature extraction methods,the modeling of insulation fault diagnosis based on modified Bayesian networks and the preliminary identification of equipment health status,as well as calculation of the posterior probability of each faulty node to achieve fault diagnosis.The method is veri-fied by experiment to have a small diagnostic error in diagnosing and predicting the insulation performance indicators of power equipment.