Statistical Parameters of Partial Discharge Used to Recognize Aged Oil-Paper Insulation

This paper presents a group of partial discharge (PD) statistical parameters for recognizing the aging degree of oil-immersed-paper insulation. These parameters are extracted by performing factor analysis on the well-known statistical operators and Weibull parameters of 2-dimensional PD statistical distributions, which consist of the PD magnitude, pulse count, and pulse phase angle. An experimental oil-immersed-paper insulation model is designed for accelerated electrical-thermal aging experiments. PD sample data is measured during the experiment for one group of specimens, while the degree of polymerization is measured for another group of specimens. The newly extracted statistical parameters of PD statistical graphs, which are constructed with PD sample data at six different aging stages, are used for aging degree recognition along with the Fisher linear discrimination equation