Document Type : Research article


Electrical Engineering Department and Centre of Excellence in Power System Management and Control, Sharif University of Technology, Tehran 1458889694, Iran


Partial Discharge (PD) measurement is one of the best solution for condition assessment of Gas Insulated Switchgears (GIS). For having Condition-based maintenance of GIS, online PD monitoring is in great importance. For this aim, Ultra High Frequency (UHF) PD sensors should be installed inside the GIS during the installation. However, in most installed GIS in industries, the internal UHF PD sensors are not installed. In this paper, a new method for on-line defect type recognition according to external UHF PD sensors and based on time-frequency representation of PD signal is proposed. In this case four artificial defect type named protrusion on main conductor, protrusion on enclosure, free moving metal particle and metal particle on spacer are implanted inside the 132 kV L-Shaped structure of one phase in enclosure GIS. The signal energy at each level of decomposed signal by Discrete Wavelet Transform (DWT) is applied for features of each defect type. The trends of signal energy variations at each frequency ranges of signal are applied for discriminating between each defect type. The Deep Feed Forward Network (DFFN) classifier is applied for PD pattern recognition. The results show the beneficial and simplicity of the proposed method for PD signal classification, independent from position of PD sensor, especially in case of online PD monitoring of GIS.


  • An online partial discharge pattern recognition method is presented based on measured Partial Discharge (PD) data from an external UHF PD sensor for condition assessment of Gas Insulated Switchgears (GIS).
  • The time-frequency representation of signal from discrete wavelet transform is applied for feature extractions of each PD defect model.
  • The feature extracted based on signal energy at each level of DWT decomposition are independent of the positioning of UHF PD sensors.
  • By using the Deep Feed-Forward Network, the classification accuracy of the proposed method for PD pattern recognition is about 94.5 %.


Main Subjects