In order to guide the state maintenance of GIS equipment more effectively

on the basis of FDTD simulation and multiple source clustering method

we diagnosed partial discharge（PD） detection in insulating spacer of 1 100 kV gas insulated switchgear（GIS）.First

the PD detection and location under live and plant were performed

and the finite difference time domain（FDTD） simulation model corresponding to the plant test structure was established.Combined with the propagation characteristics of UHF electromagnetic wave simulation

the multiple PD sources location technology were discussed.Secondly

the clustering analysis of UHF signal source of the insulation spacer was carried out based on feature optimization extraction of cumulative energy function.Finally

the regression analysis of GIS experiment platform was used to analyze the detection difference of insulation defect

which was employed to verify the live and plant test results.The test results show that the UHF pattern of this GIS insulation spacer presents as void defects

however

there is no abnormality under the pulased current method.The PD classification analysis and location results show that there are two kinds of UHF signal sources in the insulation spacer.Combined with the simulation location method

the error of the conventional location calculation method can be reduced effectively.The regression analysis shows that the sensitivity of the UHF signal is higher than that of the pulsed current method for void defect.The test sensitivity of pulsed current method will decrease gradually with the increment of the capacitance of the test object

which may be the main reason leading to the difference between the UHF signal and the pulse current method in this insulation spacer detection.The research results can provide a reference for the partial discharge detection of GIS.