Abstract
The performance and reliability of power apparatus are affected by the degradation of electrical insulation, often caused by voids in resin‐molded components. To address this issue, we investigated the use of machine learning to estimate the degree of degradation due to void discharges in epoxy resin. We prepared samples with artificial void defects and collected partial discharge data up to the dielectric breakdown, applying long‐term AC voltage. Utilizing this data, a machine learning model was developed to estimate degradation levels. The results showed that the accuracy was 55% and the false estimation rate which underestimates the degradation level was as high as 31%, which was a significant problem for practical use. Therefore, for regions of data that were difficult to classify, we constructed a weighted SVM model that predicted higher levels of degradation by weighting the data according to their degradation labels. With this approach, the accuracy remained at 57%, but the rate of underestimation of the degree of degradation was reduced to only 4%. Consequently, we were able to develop an effective estimation model that is practical for the maintenance of power apparatus. © 2023 Institute of Electrical Engineer of Japan and Wiley Periodicals LLC.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
More From: IEEJ Transactions on Electrical and Electronic Engineering
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.