基于相似度投票的牵引供电系统CT隐藏故障在线诊断
电流互感器(CT)的隐藏故障特征十分微弱,影响计量及保护设备性能,在系统出现故障或压力时容易导致系统连锁故障甚至崩溃,为及时发现CT隐藏故障并预警,需要对其进行在线诊断。现有文献的故障诊断方法均需要提取故障后特征信息,不适用于隐藏故障分析。基于此,首先根据基尔霍夫电流定律联系牵引供电系统拓扑中各支路 CT之间的电流约束关系建立关联矩阵,得到各CT解析电流值;再利用解析值与实测值相似系数受隐藏故障的影响规律建立投票规则实现CT隐藏故障在线诊断。本方法无需额外增加测量设备,仿真结果不受牵引供电系统负荷电流变化大的影响,同时对投票中的阈值选取要求不高。...
        Saved in:
      
    
          | Published in | 电力系统保护与控制 Vol. 41; no. 23; pp. 126 - 131 | 
|---|---|
| Main Author | |
| Format | Journal Article | 
| Language | Chinese | 
| Published | 
            四川省电力公司建设管理中心,四川 成都 610021%西南交通大学电气工程学院,四川 成都,610031
    
        2013
     国网成都供电公司,四川 成都 610041%西南交通大学电气工程学院,四川 成都 610031 西南交通大学电气工程学院,四川 成都 610031  | 
| Subjects | |
| Online Access | Get full text | 
| ISSN | 1674-3415 | 
Cover
| Summary: | 电流互感器(CT)的隐藏故障特征十分微弱,影响计量及保护设备性能,在系统出现故障或压力时容易导致系统连锁故障甚至崩溃,为及时发现CT隐藏故障并预警,需要对其进行在线诊断。现有文献的故障诊断方法均需要提取故障后特征信息,不适用于隐藏故障分析。基于此,首先根据基尔霍夫电流定律联系牵引供电系统拓扑中各支路 CT之间的电流约束关系建立关联矩阵,得到各CT解析电流值;再利用解析值与实测值相似系数受隐藏故障的影响规律建立投票规则实现CT隐藏故障在线诊断。本方法无需额外增加测量设备,仿真结果不受牵引供电系统负荷电流变化大的影响,同时对投票中的阈值选取要求不高。 | 
|---|---|
| Bibliography: | ZHANG Yu-ge1, 2, CHEN Shuang1, 3, HAN Zheng-qing1, GAO Shi-bin1 (1. School of Electrical Engineering, Southwest Jiaotong University, Chengdu 610031, China; 2. State Grid Chengdu Power Supply Company, Chengdu 610041, China; 3. Construction Management Center of Sichuan Electric Power Corporation, Chengdu 610021, China) 41-1401/TM traction power supply system; transformer; hidden fault; associated matrix; similarity vote; online diagnosis The hidden fault characteristics of the current transformer (CT) are very weak, which affects the performance of measurement and protection equipment and easily leads to a cascading failure or even crash when the system has fault or pressure. It is necessary to make online diagnosis for timely detection of CT hidden fault and pre-warning. The fault diagnosis methods of the existing literature require feature extraction after a fault occurs, which are not applicable to the hidden fault analysis. According to the Kirchhoff's current law, based on the current constraint relationship  | 
| ISSN: | 1674-3415 |