Research of mechanical state diagnosis techniques in GIS bus connector based on mechanical vibration

Mechanical defects are one of the main reasons of GIS equipment failure. However, there are no available detection devices and mechanical defects assessment methods for GIS mechanical fault diagnosis. Firstly, the mechanism of mechanical vibration of GIS is deeply studied in this paper. On the basis of the research on the source and detection method of GIS vibration signal, the vibration acceleration sensor is used as the sensor to develop the GIS vibration signal detection device. Secondly, the GIS mechanical vibration test platform is built in the laboratory, and the spring contact loose defect of the GIS bus connector is set up on the test platform. Then the vibration test of the GIS shell was carried out on the condition that the bus connector is normal and the connection of connector is loose using the self-developed vibration signal detection device. Finally, the vibration signal characteristics of normal and loose defect are extracted based on the spectrum analysis method. The connection state evaluation method and judgment basis of GIS bus connector are given. The result indicates that obvious vibration signal can be detected in GIS shell near insulator. In normal operation, the vibration signals in all directions to the GIS shell have the same characteristics that obvious vibration signals at 100Hz frequency as well as in the frequency range of 1000Hz-2000Hz can be found. In particular, the amplitude of vibration signal at 1400Hz is almost equal to that at 100Hz. When GIS bus connector has a loose contact, the amplitude of vibration signal is relative high and closely related to direction. Take the amplitude of 100Hz as the reference value, the ratio of amplitudes between 1400Hz and 100Hz can be regarded as the characteristic quantity indicating the connection state of bus connector. The connector can be convinced of loose contact while the ratio is larger than 2 or smaller than 0.5 inferring from the experimental data in laboratory.