METHOD FOR PREDICTING FAILURE OF ELECTRIC INSULATOR USING EVENT-DATE ANALYSIS TECHNIQUE

摘要

PURPOSE: A method for predicting failure of an electric insulator using an event-date analysis technique is provided to reduce maintenance costs of power equipments and improve supply reliability through high quality power supply. CONSTITUTION: A time is set(S8). A neutral current is obtained(S9). An Rms is calculated relative to a native signal and a high frequency through a high frequency filter(S10). A Fourier transform is performed relative to a native signal which doesn't pass the high frequency filter(S11). A G-score is calculated relative to a high frequency component and the Rms(S12). An event is calculated by using a value obtained in S12(S13). The event obtained the above step is compared with a reference number of time(S14). When the event is over the reference number of time, an even-date percentage is calculated(S15). It is determined whether the event-date percentage is over 30-40%(S16). When the event-date percentage is over 30-40%, or the event-date percentage is increased, a failure is displayed or an alarm goes off(S17).