基于日盲紫外和脉冲电流信号的异常放电融合诊断方法  

Integrated Diagnosis of Abnormal Discharge Utilizing Solar-blind Ultraviolet and Pulse Current Signals

作　　者：段生江 项恩新 陈文良 商经锐 叶超奇 陆得群 任明[3] DUAN Shengjiang;XIANG Enxin;CHEN Wenliang;SHANG Jingrui;YE Chaoqi;LU Dequn;REN Ming(Dehong Electric Power Supply Company,State Grid Yunnan Electric Power Co.,Ltd.,Dehong 678400,China;Electric Power Research Institute,State Grid Yunnan Electric Power Co.,Ltd.,Kunming 650217,China;State Key Laboratory of Electrical Insulation and Power Equipment,Xi′an Jiaotong University,Xi′an 710049,China)

机构地区：[1]云南电网有限责任公司德宏供电局,云南德宏678400 [2]云南电网有限责任公司电力科学研究院,云南昆明650217 [3]西安交通大学电工材料电气绝缘全国重点实验室,陕西西安710049

出　　处：《电工技术》2024年第21期204-207,共4页Electric Engineering

基　　金：南方电网有限责任公司科技项目(编号YNKJX M20220132)。

摘　　要：异常放电是绝缘劣化和设备故障的主要原因,不同放电检测方法各有优缺点。提出一种基于日盲紫外和脉冲电流信号的异常放电融合诊断方法。首先,在实验室环境下制作典型绝缘缺陷模型,利用高频线圈和日盲紫外传感器采集异常放电信号;然后,通过相位分析图谱提取放电特征,构建包含放电相位信息的数据集;最后,使用神经网络构建放电类型识别模型,与不同数据源的异常放电诊断效果进行对比。结果表明,基于改进反向传播神经网络模型的识别准确率可达96.4%;相比于单一检测数据,使用光电融合数据诊断效果更优。Abnormal discharge is the main cause of insulation degradation and equipment failure,and each independent detection method has specific pros and cons.This work studied an integrated method of diagnosing abnormal discharge by simultaneously utilizing signals of solar-blind ultraviolet and pulse current.First the actual models of typical insulation defects were prepared in the laboratory,and the abnormal discharge signals were collected through high-frequency current transformer and solar-blind ultraviolet sensor.Then the features of abnormal discharge were extracted through phase-resolved partial discharge pattern,and the dataset containing phase information of discharge was obtained.Finally a discharge type identification model was established using a neural network,and a comparison of abnormal discharge diagnosis performance was conducted using different data sources.The results showed that the identification accuracy of the modified back propagation neural network model could reach 96.4%.Compared with single-source detection data,the use of optical-electrical integrated data achieved superior diagnosis performance.

关 键 词：异常放电 融合诊断 特征提取 类型识别 

分 类 号：TM855[电气工程—高电压与绝缘技术]