Assessment of the Insulation Status Aging in Power Cable Joints Using Support Vector Machine

This study presents a method for predicting the insulation status of power cable joints based on partial discharges (PDs). First, PD data are collected from aging experiments with 13 power cable joints involving artificial defects. Second, two PD features with a high identification of aging status are extracted from 104 PD features. Additionally, the initial and final stages of insulation aging are defined, and the labeled PD data are trained using a support vector machine (SVM). Finally, a method to detect insulation failure is proposed by estimating the minimum distance from the PD to the hyperplane of the SVM using nonlinear programming. The alert is defined as the turning of the curve in the final stage. The proposed method verifies that the warning and danger alerts are issued before insulation failure occurs for each sample.