基于TWFPN算法的配电网故障主动诊断研究  

Research on the active fault diagnosis of distribution network based on TWFPN algorithm

作　　者：张婷婷[1] 张信群[1] 许敏 ZHANG Tingting;ZHANG Xinqun;XU Min(College of Mechanical and Automotive Engineering,Chuzhou Polytechnic,Chuzhou 239000,China)

机构地区：[1]滁州职业技术学院机械与汽车工程学院,安徽滁州239000

出　　处：《河南工程学院学报（自然科学版）》2022年第1期55-60,共6页Journal of Henan University of Engineering：Natural Science Edition

基　　金：安徽省质量工程教学研究重大项目(2018jyxm0669);安徽省高校优秀青年人才支持计划重点项目(gxygZD2021157)。

摘　　要：针对传统矩阵计算对闭环有源配电网故障诊断存在的逻辑简单、维护困难问题,基于闭环有源配电网时序规则与加权模糊Petri网(WFPN),设计了一种时序连接加权模糊Petri网(TWFPN)故障诊断模型。首先,引入时序连接,防止对配电网故障线路误判;然后,采用高斯分布函数集中预处理SOE信息,避免信息丢失或拒动;最后,通过实例验证了基于TWFPN的配电网故障诊断模型可有效诊断当前广泛应用的有源配电网单线路故障和多线路故障,诊断结果与实际故障情况相符,且相较于典型的配电网故障诊断算法,该模型的诊断正确率可达到95%。In view of the problems of simple logic and difficult maintenance for the fault diagnosis of closed-loop active distribution network by traditional matrix calculation,this paper designs a time series connection weighted fuzzy Petri net(TWFPN)fault diagnosis model based on the time sequence rules and weighted fuzzy Petri net(WFPN).By introducing the time series connection,the misjudgment of the fault line of distribution network is prevented;the SOE information is preprocessed by Gaussian distribution function to avoid information loss or rejection.Finally,the case study shows that the fault diagnosis model based on TWFPN can effectively diagnose single line fault and multi-line fault of active distribution network,and the diagnosis results are in accordance with the actual fault situation.Compared with the typical fault diagnosis algorithm,the diagnosis accuracy of this model can reach 95%.

关 键 词：配电网 TWFPN算法 故障诊断 

分 类 号：TM712[电气工程—电力系统及自动化]