Self-Normalizing Multivariate Analysis of Polymer Insulator Leakage Current Under Severe Fog Conditions

Self-normalizing multivariate (SNM) analysis method was proposed to reflect characteristics of leakage current for monitoring dynamic surface discharges to flashover on polymer insulator in severe fog-polluted conditions. Based on the observation of discharge phenomena and temporal measurement of leakage current, the process just prior to flashover was separated into four stages: dominated resistive current, partial discharge, stable dry-band discharge, and pre-flashover. With discharge propagation in these stages, the SNM map was obtained. With increasing the pollution level, the SNM map shows no obvious difference in the resistive current dominated stage, while it looses more symmetry in both stable dry-band discharge and pre-flashover stages, which indicates the highest probability of surface flashover occurrence under the most severe fog condition investigated.