变电站故障诊断的时间不确定性问题研究  

Research on temporal uncertainty of substation fault diagnosis

作　　者：佘维[1] 郭金叶[1] 李江林 姬波[1] 叶阳东[1] 

机构地区：[1]郑州大学信息工程学院,河南郑州450000 [2]许继电气股份有限公司,河南许昌461000

出　　处：《电力系统保护与控制》2012年第8期1-9,共9页Power System Protection and Control

基　　金：国家自然科学基金(60773048);国家重点实验室开放课题基金(RCS2009K003);湖南省教育厅科学研究课题(09C1257)~~

摘　　要：针对变电站故障诊断的时间不确定性,提出一类带时间隶属函数的模糊Petri网(TM-FPN)用于构建变电站故障诊断模型。该模型利用模糊时间片和加权数学期望计算故障信号的时间分布,并采用一种加权方差分析信号分组的离散程度,选取离散程度最低的分组计算其时间隶属度,进而使用加权模糊推理进行诊断。仿真实验表明,该模型能有效地处理变电站故障诊断中时间不确定性问题,并能定量地分析信号时间分布离散度、时间差值函数以及重复出现的信号对诊断结果的影响。According to temporal uncertainty of substation fault diagnosis, a fuzzy Petri nets with temporal membership（TM-FPN） is proposed to construct the fault diagnosis model of substation. Fuzzy timestamp and weighted mathematical expectation are used to calculate temporal distribution of fault signals. Then, a kind of weighted variance is applied to analyze the discrete degree of signal groups. The next is that calculating the temporal membership of the lowest discrete degree group. Finally, this paper adopts weighted fuzzy reasoning to carry out fault diagnosis. The simulation experiment indicates that TM-FPN can effectively deal with temporal uncertainty of substation fault diagnosis, and can be used to analyze quantitatively the influence of temporal distribution, temporal interval and repeated signals on fault diagnosis results.

关 键 词：变电站 故障诊断 PETRI网 时间不确定性 时间隶属度 

分 类 号：TM73[电气工程—电力系统及自动化]