Identification of the Aging Stage of Transformer Oil-Paper Insulation via Raman Spectroscopic Characteristics
Dingkun Yang, Weigen Chen, Fu Wan, Yongkuo Zhou, Jiaxuan Wang
Abstract
Diagnosis of the aging stage of oil-paper insulation is an important technical means for preventing aging failure of oil-paper insulation equipment. This article proposes a Raman spectroscopy diagnostic model to assess the aging status of transformer oilpaper insulation by analyzing the Raman spectra of the insulating oil. Oil-paper samples in different aging stages were obtained from electrical and thermal accelerated aging tests in the lab. The Raman spectra of insulating oil were collected in the range from 400 to 3000 cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-1</sup> , which covers the Raman signals of the aging products of oilpaper insulation. Spectral peak contour features were extracted based on Voigt fitting of the Raman spectra and studied using the grid search optimized one-against-one multi-class support vector machine (OAO MCSVM) for diagnosis of the aging stage. The results show that the diagnostic accuracy of the diagnostic model based on the spectral peak contour features was 92% (46/50). Spectral peak contour features demonstrate an ability to reflect the aging process, and the proposed diagnostic method shows potential for monitoring oil-paper insulation.