Research on a General Fast Analysis Algorithm Model for Pd Acoustic Detection System: The Algorithm Model Design and Its Application

2019 International Conference on Smart Grid and Electrical Automation (ICSGEA) Pub Date : 2019-08-01 DOI:10.1109/icsgea.2019.00014

W. Si, C. Fu, K. Gao, Jia-Min Zhang, Lin He, Hailong Bao, Xin-ye Wu

{"title":"Research on a General Fast Analysis Algorithm Model for Pd Acoustic Detection System: The Algorithm Model Design and Its Application","authors":"W. Si, C. Fu, K. Gao, Jia-Min Zhang, Lin He, Hailong Bao, Xin-ye Wu","doi":"10.1109/icsgea.2019.00014","DOIUrl":null,"url":null,"abstract":"Nowadays, the detection of acoustical emission is widely used for fault diagnosis of gas insulated substations (GIS) in normal operation and factory tests, which is called 'non-conventional' method recommended in the standard IEC TS 62478-2016 and GIGRE D1.33 444. In this paper, to develop a data analyzer for acoustic detection (AD) system to make an assistant diagnosis for technical personnel or equipment operation and maintenance personnel, based on the previous research on the experimental research, pattern identification with phase compensation and the software development, the algorithm model design and its application is given in detail. For the acoustical emission signals (n, ti, qi), the BP artificial neural network optimized by genetic algorithm (GA-BP) is used as a classifier based on the fingerprint consisting of several statistic operators, which are derivate form typical 2D histograms of PRPD with identification with phase compensation (IPC). Experimental results show that the comprehensive algorithm model designed for identification is practical and effective.","PeriodicalId":201721,"journal":{"name":"2019 International Conference on Smart Grid and Electrical Automation (ICSGEA)","volume":"87 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 International Conference on Smart Grid and Electrical Automation (ICSGEA)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/icsgea.2019.00014","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 2

Abstract

Nowadays, the detection of acoustical emission is widely used for fault diagnosis of gas insulated substations (GIS) in normal operation and factory tests, which is called 'non-conventional' method recommended in the standard IEC TS 62478-2016 and GIGRE D1.33 444. In this paper, to develop a data analyzer for acoustic detection (AD) system to make an assistant diagnosis for technical personnel or equipment operation and maintenance personnel, based on the previous research on the experimental research, pattern identification with phase compensation and the software development, the algorithm model design and its application is given in detail. For the acoustical emission signals (n, ti, qi), the BP artificial neural network optimized by genetic algorithm (GA-BP) is used as a classifier based on the fingerprint consisting of several statistic operators, which are derivate form typical 2D histograms of PRPD with identification with phase compensation (IPC). Experimental results show that the comprehensive algorithm model designed for identification is practical and effective.

查看原文

微信好友朋友圈 QQ好友复制链接

本刊更多论文

Pd声探测系统通用快速分析算法模型研究:算法模型设计及应用

目前，声发射检测被广泛应用于气体绝缘变电站(GIS)的正常运行和工厂试验中的故障诊断，在标准IEC TS 62478-2016和GIGRE D1.33 444中被推荐为“非常规”方法。为了开发一种用于声学检测(AD)系统的数据分析仪，为技术人员或设备运维人员进行辅助诊断，本文在前人实验研究、相位补偿模式识别和软件开发的基础上，详细介绍了算法模型设计及其应用。对于声发射信号(n, ti, qi)，采用遗传算法优化的BP人工神经网络(GA-BP)作为基于指纹的分类器，该指纹由几种统计算子组成，这些统计算子是典型的PRPD二维直方图的衍生，具有相位补偿识别(IPC)。实验结果表明，所设计的综合算法模型是实用有效的。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文去求助

来源期刊

2019 International Conference on Smart Grid and Electrical Automation (ICSGEA)

自引率

0.00%

发文量