Typical Interference Patterns Happened To GIS Partial Discharge Detection By Using A UHF Approach

The present research aims to establish a pattern base of interference signals for UHF-based detection of GIS partial discharge so that discharge signals can be effectively distinguished from interference signals during GIS partial discharge detection. To this end, an experimental platform was established for the study of interference patterns of GIS partial discharge, where both external UHF sensor and dialectic-windowed UHF sensor were applied to assess the five types of typical interferences from mobile phone signals (both GSM and CDMA), corona discharge, suspensor discharge, surface discharge and light signals respectively. Experimental results were analyzed and presented through scatter plot, histogram, three-dimensional plot, gray scale and Time-Frequency plot to illustrate the five kinds of typical interference patterns. The research findings conclude that:1) Five typical interference patterns differ from one another, which could be utilized as criteria in identifying the inference pattern; 2) the two different UHF sensors identify different patterns of interference given the same external conditions; 3) dialectic-windowed UHF sensor outperforms external UHF sensor in terms of anti-interference capability; 4) UHF methods show little sensitivity in detecting corona discharge signals.