Identification Method of Abnormal Contact Defect on GIS Conductor Base and Disconnector Contact Based on Ensemble Empirical Mode Decomposition

Gas insulated switchgear (GIS) is widely used in power system because of its high reliability and compact structure. Mechanical defect is one of the main reasons for GIS equipment abnormal vibration, even insulation aging breakdown and gas leakage. In order to detect and identify the internal mechanical fault of GIS, this paper first developed a detection system of GIS vibration signal based on piezoelectric acceleration sensor. After then, 110kV-3150A GIS vibration defect simulation test platform was built in laboratory, and two kinds of internal mechanical defects such as loose bolts of conductor base and poor contact of disconnecting switch were set. Meanwhile, vibration fault signals under different current were detected. Finally, based on the ensemble empirical mode decomposition (EEMD) is used to decompose the fault signals of defects. The IMF components with large correlation coefficient were extracted as the screening signals, and the time-frequency matrix is transformed to extract the abnormal features of fault signals. The Gustafson-Kessel (G-K) fuzzy clustering algorithm is used to classify and verify the fault signals of GIS. The results show that the two kinds of defects present the difference of signal amplitude in different time-frequency areas. The GIS mechanical fault diagnosis method based on the set empirical mode decomposition and G-K fuzzy clustering in time-frequency domain can effectively extract the fault features and accurately realize the fault classification and recognition.