基于马尔可夫转移场的SF_(6)分解气体紫外吸收光谱检测  

Detection of SF_(6)decomposition gases using ultraviolet absorption spectroscopy based on Markov transition fields

作　　者：王秀君 罗林[2] 于淼[3] 曹宇鹏 WANG Xiujun;LUO Lin;YU Miao;CAO Yupeng(China Telecom Corporation Limited Zhejiang Branch,Hangzhou 310001,China;School of Information and Control Engineering,Liaoning Petrochemical University,Fushun,Liaoning 113001,China;College of Electrical Engineering,Zhejiang University,Hangzhou 310027,China;State Grid Jibei Electric Power Co.,Ltd.,Beijing 100054,China)

机构地区：[1]中国电信集团有限公司浙江分公司,杭州310001 [2]辽宁石油化工大学信息与控制工程学院,辽宁抚顺113001 [3]浙江大学电气工程学院,杭州310027 [4]国网冀北电力有限公司,北京100054

出　　处：《浙江电力》2025年第3期116-124,共9页Zhejiang Electric Power

基　　金：国家自然科学基金青年项目资助(61703191);工业控制技术国家重点实验室开放课题资助(ICT202B41)。

摘　　要：气体绝缘开关设备绝缘故障引起的局部放电会导致绝缘介质SF_(6)气体分解,分解产生的氟硫化物会与设备中的水分和氧气发生反应,生成SO_(2)、CS_(2)和H_(2)S等化合物,这3种特征气体的浓度能够反映绝缘故障的类型和严重程度。对此,提出一种基于MFT(马尔可夫转移场)和深度学习的方法,用于分析特征气体的紫外光谱图,识别具有重叠吸收峰的痕量气体混合物。该方法首先将一维序列数据转化为二维图像,使混合气体的紫外光谱数据以图像形式呈现,从而使其特征信息更加直观。随后,将这些图像输入至深度学习模型中进行识别分类。实验结果表明,与其他模型相比,该方法具有更高的识别准确率。Partial discharge caused by insulation faults in gas-insulated switchgear can lead to the decomposition of the insulating medium,SF_(6)gas.The byproducts of this decomposition,fluorosulfides,can react with moisture and oxygen in the equipment,generating compounds like SO_(2),CS_(2),and H_(2)S.The concentrations of these three charac⁃teristic gases can reflect the type and severity of insulation faults in the equipment.To address this,the paper pro⁃poses a method based on Markov transfer fields(MTF)and deep learning to analyze the ultraviolet spectra of these characteristic gases and identify trace gas mixtures with overlapping absorption peaks.This method first converts one-dimensional sequence data into two-dimensional images,presenting the ultraviolet spectral data of the mixed gases in a more intuitive image format.Subsequently,these images are input into a deep learning model for classifi⁃cation and recognition.Experimental results demonstrate that this method achieves a higher recognition accuracy compared to other models.

关 键 词：紫外吸收光谱法 马尔可夫转移场 深度学习 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]