Defect type identifying method for GIS oscillation impulse voltage withstanding test

A defect type identifying method for a GIS oscillation impulse voltage withstanding test forms a pulse waveform-time series by extracting a single pulse time-domain waveform and a corresponding discharge time by using a custom pulse extraction window; uses an equivalent periodic folding algorithm to form an equivalent phase-peak distribution chart according to the pulse waveform-time series; uses a statistical operator to extract feature parameters of the equivalent phase-peak distribution chart to form a discharge fingerprint, and processes the discharge fingerprints of different discharge defects to form a discharge fingerprint database; uses a BP neural network to train the discharge fingerprint database to form a discriminant function F; and uses the discriminant function F to calculate the discharge fingerprint of a defective GIS specimen to be discriminated, and identifies the type of a defect. The method can effectively identify the type of defects of partial discharge under an oscillation impulse volt