基于多源监测数据挖掘的电力设备状态诊断State Diagnosis of Power Devices Based on Multi-source Monitoring Data Mining
郑一鸣,孙翔
ZHENG Yiming,SUN Xiang
摘要(Abstract)：
随着设备状态检修及信息化建设的发展,输变电设备的状态评估和诊断成为电力系统安全稳定运行的关键技术。传统设备状态诊断多是基于离线检测参量或单一监测参量进行的,在此基础上提出一种基于多源监测数据的综合分析方法,提供了利用多源监测结果分析设备状态和缺陷的思路。利用该方法对一台带缺陷特高压高抗的多源监测数据进行时序及相关性分析,根据分析结果对设备的缺陷溯源。应用结果表明,该方法可应用于电力设备的缺陷诊断,并有效提高了诊断的准确性。
With the development of condition-based power device maintenance and information construction,state assessment and diagnosis of power devices in transmission and transformation systems have increasingly become the key technology to ensure operation safety and stability of the power system. The traditional state diagnosis of power devices is mostly based on offline monitoring parameters or single monitoring parameter. In this paper, a comprehensive analysis method is presented based on the mining of multi-source monitoring data and a new thought of device state and defect analysis using multi-source monitoring result is provided. This method is applied to analyze the time sequence and correlation of the multi-source monitoring data of a defective UHV reactor. According to the analysis results, the device defects are investigated. The application practice shows that the method can be applied to defect diagnosis of power devices and can effectively improve diagnostic accuracy.
关键词(KeyWords)：
电力设备;在线监测;数据挖掘;数据清洗;时序分析;相关性分析;状态诊断
power device;on-line monitoring;data mining;data cleaning;time sequence analysis;correlation analysis;state diagnosis
基金项目(Foundation): 浙江省电力公司科技项目(5211DS150026)
作者(Author):
郑一鸣,孙翔
ZHENG Yiming,SUN Xiang
DOI: 10.19585/j.zjdl.2016.05.001
参考文献(References)：
- [1]孙才新.输变电设备状态在线监测与诊断技术现状和前景[J].中国电力,2005,38(2):1-7.
- [2]刘黎,何文林,刘岩,等.输变电设备状态在线监测与故障诊断系统分析软件设计[J].计算机系统应用,2011,20(8):27-32.
- [3]郭碧红,杨晓洪.我国电力设备在线监测技术的开发应用状况分析[J].电网技术,2004,23(8):65-68.
- [4]王德文,阎春雨.变电站在线监测系统的一体化建模与模型维护[J].电力系统自动化,2013,37(23):78-82.
- [5]宁辽逸,吴文传,张伯明.运行风险评估中的变压器时变停运模型(一)基于运行工况的变压器内部潜伏性故障的故障率估计方法[J].电力系统自动化,2010,34(15):9-13.
- [6]郭创新,高振兴,刘毅,等.采用分层多源信息融合的电网故障诊断方法[J].高电压技术,2010(12):2976-2983.
- [7]严英杰,盛戈皞,陈玉峰,等.基于时间序列分析的输变电设备状态大数据清洗方法[J].电力系统自动化,2015,39(7):138-144.
- [8]魏武雄.时间序列分析:单变量和多变量方法[M].北京:中国人民大学出版社,2009.
- [9]孙翔,何文林,邱炜,等.基于显著性差异的油浸倒置式电流互感器氢气阈值分析[J].浙江电力,2014,33(6):20-25.
- [10]陈燕娜.生产管理系统在供电企业的应用探究[J].现代制造,2014,(27):163-164.
- [11]DL/T 722-2014变压器油中溶解气体分析和判断导则[S].北京:中国电力出版社,2014.
- [12]GB/Z 24846-2009 1 000 k V交流电气设备预防性试验规程[S].北京:中国电力出版社,2009.
- [13]刘文艺,汤宝平,蒋永华.一种自适应小波消噪方法[J].振动、测试与诊断,2011,31(1):74-77.
- [14]XIONG H,WU J,CHEN J.Kmeans clustering versus validation measures a data distribution perspective[C]//In KDD.2006:318-331.