电工技术学报2018，Vol.33Issue(5)：1133-1142,10.DOI:10.19595/j.cnki.1000-6753.tces.162019
油纸绝缘老化拉曼光谱特征量提取及诊断方法
The Raman Spectral Feature Extraction and Diagnosis of Oil-Paper Insulation Ageing
摘要
Abstract
Accurate diagnosis of oil-paper insulation ageing stage serves as an important technology to prevent major accidents of oil-paper insulation equipment. The extraction of effective characteristics which reflect the insulation ageing proves to be the essential step. Raman spectroscopy has been demonstrated that it has great potential of mixture composition analysis and condition diagnosis. In this paper, the Raman spectral features of oil-paper insulation samples were researched and extracted based on the Raman spectroscopy platform. Firstly, the samples in different ageing stages were prepared by thermal accelerated ageing process. Then the principal component analysis method was employed to reduce dimensionality of the obtained Raman spectral data. Secondly, the spectral features strongly corresponding to the ageing of oil-paper insulation were extracted. Based on Raman spectral features, multi-classification support vector machine optimized by particle swarm algorithm was used to set up an oil-paper insulation ageing diagnosis model. Finally, the diagnostic capability and universality of the established algorithm were verified by the samples made in lab.关键词
拉曼光谱/油纸绝缘/特征量/老化诊断Key words
Raman spectroscopy/oil-paper insulation/features/ageing diagnosis分类
信息技术与安全科学引用本文复制引用
邹经鑫,陈伟根,万福,范舟..油纸绝缘老化拉曼光谱特征量提取及诊断方法[J].电工技术学报,2018,33(5):1133-1142,10.基金项目
国家自然科学基金青年基金(61605020)和高等学校学科创新引智计划(111计划)(B08036)资助项目. （61605020）
