To diagnose and evaluate the gas-insulated metal-enclosed switchgear（GIS） operating state by analyzing the SF

6

decomposition products

it is necessary to find out the appropriate characteristic parameters and establish an association rule between the decomposed gas composition and GIS insulation deterioration. Consequently

we experimentally studied the long-term sustainable discharge decomposition by needle-plate defect model in order to simulate the whole process of GIS insulation d

eterioration. SO

2

F

2

SOF

2

SO

2

CO

2

and CF

4

were selected as feature components

their concentration trend curves at different stages were captured. SO

2

SOF

2

and SO

2

F

2

showed a trend of "Logistic" curve

which displayed a satisfied consistence. Concentration of SO

2

rate of（SO

2

F

2

+SOF

2

+SO

2

） and the ratio of φ（SO

2

F

2

+SOF

2

+SO

2

）/φ（CO

2

+CF

4

） were proposed as characteristic parameters and their physical significance were also explained. Three characteristic parameters change significantly at different stages.Moreover

fuzzy analysis was used to identify the severity of discharge

then corresponding fuzzy judgment model was established

and GIS insulation deterioration trend was divided into three states of "normal stage"

"deterioration stage"

and "saturation stage". The classification results are consistent with the experimental ones.