A Power System Fault Diagnosis Method based on Mutual-Information Network

Information theory is used to study and deal with the uncertainty existing in the power system fault process, and a fault diagnosis method based on mutual-information network is proposed. The information of protective relays and circuit breakers received by SCADA is preprocessed, the elements in fault areas are taken as the input attributes of mutual-information network, and the fault information is taken as the target attributes to test the association between fault and omen. The results of tests have justified that this method can recognize the fault elements effectively, and give the association degree between state of elements and fault information.