Analysis of phase-resolved partial discharge patterns of voids based on a stochastic process approach

A method is presented for the determination of physical discharge parameters for partial discharges (PDs) of voids in solid insulation. Based on a recently developed stochastic theory of PD processes, a statistical analysis of a measured phase-resolved partial discharge (PRPD) pattern allows the determination of the relevant physical parameters like first electron availability or decay time constants for deployed charge carriers. These parameters can be estimated directly from the measured patterns without the need of performing simulations. Furthermore, error bounds for the parameter values can be given. The parameter estimation algorithm is based on the analysis of a contiguous region of the PRPD pattern where this region can be chosen nearly arbitrarily. Thus, even patterns with several active PD defects or patterns which are corrupted by noise can be analysed. The method is applied to a sequence of patterns of a void in epoxy resin. The change in first electron availability in the course of a day can be determined quantitatively from the data while the other physical parameters remain constant.