Method for suppressing random pulse interference of partial discharge based on fuzzy clustering

In order to suppress random pulse interference in field detection of partial discharge in power equipments, pulse features is extracted using equivalent time-frequency analysis. Based on adaptive fuzzy clustering of the features, concentration degree of phases is used as the indicator to distinguish partial discharge signal from pulse interference. The results of field data analysis indicate that our method can effectively remove random pulse interference in partial discharge signal. This study provides good theoretical and practical basis for insulation state assessment in power equipments.