GIS partial discharge fault discrimination method based on feature similarity

A GIS partial discharge fault discrimination method based on feature similarity adopts a VMD algorithm to preprocess a signal, has stronger adaptive capability compared with wavelet analysis, does not need to manually determine a wavelet basis function, reduces the influence on signal features, and meanwhile, as an improved EMD algorithm, the VMD algorithm is stronger in anti-aliasing capability, fully considers the narrowband property, and improves the reliability of the EMD algorithm. The signal-to-noise ratio of the components obtained through decomposition is higher, and each analytic signal has more physical significance. And meanwhile, the energy information after VMD decomposition and the original signal power spectral density information are adopted to judge whether a fault exists or not, diagnosis is performed after the fault is determined, and compared with direct fault identification, calculation can be simplified, and the system efficiency is improved. The traditional discharge frequency, mean val