Aiming at the frequent occurrence of GIS faults

a fault detection and analysis method based on vibrationcharacteristic estimation is proposed. Firstly

the vibration signal characteristic of the metal enclosures of GIS is ex-tracted and the state-space model of the system is established based on the general auto-regressive model. Then

theparticle filter algorithm is hired to estimate the vibration characteristic of GIS. At the same time

the absolute valueof the residual between the characteristic estimate and the actual measurement is selected as the fault detection indi-cator combined with the adaptive threshold to detect the fault. Finally

the mechanical fault and the discharge faultare isolated according to the trend of the vibration characteristic residual. The simulation analysis of the measured vi-bration data before and after different faults of multiple sets of substation GIS equipment found that the vibrationcharacteristics estimation can not only accurately detect equipment faults

but also effectively isolate equipment me-chanical faults and discharge faults.