Aging condition assessment of oil-paper based on principal component and factor analysis of partial discharge

A new statistical analysis method of phase-resolved partial discharge (PRPD) is proposed for the aging condition assessment of oil-paper. The accelerated aging test of a batch of oil-paper specimens with internal flat air gap is developed and the PD signal is measured periodically. Based on the conventional PRPD pattern, a group of new features are extracted by using the method of Principal Component and Factor Analysis. These features not only are independent with each other but also all have their own physical meaning. They represent the most information of PRPD pattern by limited number of variances. The clustering results of new features extracted at different aging stages suggest that the proposed method provide a powerful approach to identify the oil-paper's aging conditions.