Aging Evaluation of Transformer Oil-Paper Insulation Using Partial Discharge Patterns and Clustering Methods

This paper presents an analysis of the aging of oil-paper insulation using partial discharge (PD) patterns and clustering methods. The four-quadrant model of partial discharge patterns is employed. The PD monitoring system is designed to detect and simulate the behavior of insulation paper in oil-immersed transformers subjected to long-term electrical stress in a laboratory environment. A high-frequency current transformer (HFCT) is used for signal measurement to evaluate the aging process. Initially, an electrical monitoring method is utilized to observe the aging of the insulation paper, and the phase-resolved pulse sequence (PRPS) method is applied to generate 3D maps based on PD sequence and amplitude, which are used to identify the form of discharge. Furthermore, the maximum peak values of signals are analyzed to assess the aging trend of the insulation paper. The four-quadrant model of PRPS patterns is then examined in relation to the four stages of the paper aging process. The relationship between the peak-to-average ratio of PD signals indicates that Quadrants I and III serve as effective bases for two clustering methods and provide a reference for future evaluations of insulation paper aging trends.