1. Oil-paper aging evaluation by fuzzy clustering and factor analysis to statistical parameters of partial discharges
- Author
-
Stanislaw Grzybowski, Jian Li, Lijun Yang, and Ruijin Liao
- Subjects
Fuzzy clustering ,business.industry ,Fuzzy set ,Statistical parameter ,Pattern recognition ,Thermal aging ,Fuzzy logic ,Kernel (statistics) ,Partial discharge ,Electronic engineering ,Artificial intelligence ,Electrical and Electronic Engineering ,business ,Cluster analysis ,Mathematics - Abstract
A thermal aging experiment was conducted for oil-paper insulation to evaluate the insulation's response under thermal stress. Oil-papers at different aging stages were used to create oil-paper bound gas cavity specimens, which were used to collect data on degrees of polymerization (DP) and partial discharge (PD). Evaluation of oil-paper aging was based on statistical operators of PD and factor analysis was used to extract the principle parameters of PD. Three types of fuzzy clustering approaches were used to classify PD of aged oil-paper: the fuzzy c-means, the kernel fuzzy c-means, and the possibilistic fuzzy c-means. The clustering results showed that the possibilistic fuzzy c-means clustering was capable of classifying PD of oil-paper bound gas cavity specimens. The factor analysis method was also verified to be helpful in fuzzy clustering of PD data samples by reducing number of parameters.
- Published
- 2010