基于拉曼光谱的电力变压器绝缘故障自动定位系统设计  

Design of automatic insulation fault location system for power transformer based on raman spectroscopy

作　　者：龙英凯 王谦 李勇 李龙 毛磊 LONG Yingkai;WANG Qian;LI Yong;LI Long;MAO Lei(Electric Power Research Institute,State Grid Chongqing Electric Power Company,Chongqing 401123,China;China Southern Power Grid Materials Co.Ltd.,Guangzhou Guangdong 510640,China)

机构地区：[1]国网重庆市电力公司电力科学研究院,重庆401123 [2]南方电网物资有限公司,广东广州510640

出　　处：《自动化与仪器仪表》2020年第1期105-108,共4页Automation & Instrumentation

基　　金：重庆市基础研究与前沿探索项目(No.cstc2018jcyjAX0068);国网重庆市电力公司科技资助项目

摘　　要：电力变压器绝缘故障定位是提高电力变压器稳定性的关键,提出一种基于拉曼光谱的电力变压器绝缘故障自动定位方法。采用卷积稀疏自编码器进行电力变压器绝缘故障特征检测,提取电力变压器绝缘故障的拉曼光谱特征量,根据电力变压器的能源、负荷差异性进行受扰响应特性分析,构建电力变压器绝缘故障辨识模型,采用深度学习方法进行电力变压器绝缘故障检测中的收敛性判断,根据拉曼光谱的频谱特征分布进行电力变压器绝缘故障的自动定位。结合API接口和SS/E23节点网络拓扑模型实现电力变压器绝缘故障自动定位系统设计。仿真测试结果表明,采用该方法进行电力变压器绝缘故障定位的准确性较高,故障的可视化分辨能力较强。Insulation fault location of power transformers is the key to improve the stability of power transformers.An automatic fault location method for power transformers based on Raman spectroscopy is proposed.The characteristic of power transformer insulation fault is detected by convolution sparse self-encoder,and the Raman spectrum characteristic of power transformer insulation fault is extracted.According to the energy source and load difference of power transformer,the characteristic of disturbance response is analyzed.The insulation fault identification model of power transformer is constructed,and the convergence of power transformer insulation fault detection is judged by using depth learning method,and the automatic location of insulation fault of power transformer is carried out according to the spectrum characteristic distribution of Raman spectrum.Combined with API interface and SS/E23 node network topology model,the design of power transformer insulation fault automatic location system is realized.The simulation results show that the proposed method is more accurate and has better visual resolution ability for power transformer insulation fault location.

关 键 词：拉曼光谱 故障检测 自动定位 辨识模型 深度学习 系统设计 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]