Air-gap discharge development characteristic research and process partition based on energy-wavelet moment in oil-paper insulation of transformer

The detection of partial discharge is of great significance in oil paper insulation and assessment of the operation status of transformers. This paper builds the air-gap discharge model in oil-paper insulation, and presents a three-dimensional spectra with considering the average discharge energy per second. Then drawing the two-dimensional gray intensity image based on discharge energy and phase. Both the global and local features of gray intensity images are obtained by introducing the wavelet theory into the extracting process of moment characteristic parameters. Finally, the wavelet moments at different times are analyzed by using fuzzy C-means clustering method, the whole process of air-gap discharge is divided as four parts: the start discharge and shock developing stage, the weak discharge stage, the outbreak discharge stage and the pre-breakdown stage.