Abstract: Partial discharge fault diagnosis and location of gas insulated metal enclosed transmission line

（GIL）is of great significance for its internal insulation protection. Because GIL is fully enclosed structure，the

existing external detection method is not sensitive to the internal defect detection，so it is unable to realize the

quantitative description and location of partial discharge. Therefore，a partial discharge fault diagnosis and location method based on Teager energy spectrum extreme learning machine was proposed. Firstly，Teager energy operator was used to extract the energy fluctuation characteristics of different discharge fault locations. Secondly，wavelet packet transform was used to calculate the proportion of characteristic frequency band of discharge fault energy at different positions. Finally，combined with extreme learning machine（ELM），the partial discharge fault of GIL equipment was located and diagnosed . The results show that the energy fluctuation of partial discharge defects at different positions is obviously different，the energy proportion of special frequency band analyzed by wavelet packet transform（WPT）can be used as the basis of fault classification. Compared with other on-line monitoring methods，ELM can effectively diagnose and locate partial discharge fault of GIL equipment.

      Key words: gas insulated metal enclosed transmission line（GIL）；partial discharge（PD）；pulse current；

Teager energy spectrum；wavelet packet transform（WPT）；fault location and diagnosis