基于故障波形时频特征配网故障识别方法研究
发布时间：2018-12-13 06:06
【摘要】：配电线路故障类型的准确识别可为线路运维人员提供方向性的指导。基于对故障波形时频特征的分析,提出了一种配电线路故障类型识别方法,通过对不同类型故障波形建立模型与理论分析,从时域、频域和电弧3个方面提出可以表征不同故障类型波形特点的特征参量;并提出依据故障波形数据提取特征参量的计算公式,将多参量融合作为依据建立识别逻辑,通过对输入波形数据特征量的检测归类,实现对不同原因引起的配网线路故障类型的自动识别。最终利用美国电力研究协会(EPRI)提供的136组现场故障波形数据对算法进行了闭环控制与验证,结果显示识别成功率达到90%,证实了利用故障波形时频特性实现配电线路故障类型识别的可行性。
[Abstract]:Accurate identification of fault types of distribution lines can provide directional guidance for line operators. Based on the analysis of time-frequency characteristics of fault waveforms, a fault type identification method for distribution lines is proposed. The characteristic parameters which can characterize the waveform characteristics of different fault types are proposed in frequency domain and arc domain. The formula of extracting characteristic parameters based on fault waveform data is put forward, and the multi-parameter fusion is taken as the basis of establishing identification logic, and the characteristic quantity of input waveform data is detected and classified. Automatic identification of fault types of distribution network caused by different reasons is realized. Finally, the closed-loop control and verification of the algorithm are carried out using 136 sets of field fault waveform data provided by the American Electric Power Research Association (EPRI). The results show that the success rate of recognition is 90%. It is proved that it is feasible to use the time-frequency characteristic of fault waveform to identify the fault type of distribution line.
【作者单位】： 上海交通大学电子信息与电气工程学院;云南电力科学研究院;国网河南省电力公司电力科学研究院;
【基金】：国家自然科学基金(51307109) 国家电网公司科技项目(SGTYHT/14-JS-188)资助
【分类号】：TM75
[Abstract]:Accurate identification of fault types of distribution lines can provide directional guidance for line operators. Based on the analysis of time-frequency characteristics of fault waveforms, a fault type identification method for distribution lines is proposed. The characteristic parameters which can characterize the waveform characteristics of different fault types are proposed in frequency domain and arc domain. The formula of extracting characteristic parameters based on fault waveform data is put forward, and the multi-parameter fusion is taken as the basis of establishing identification logic, and the characteristic quantity of input waveform data is detected and classified. Automatic identification of fault types of distribution network caused by different reasons is realized. Finally, the closed-loop control and verification of the algorithm are carried out using 136 sets of field fault waveform data provided by the American Electric Power Research Association (EPRI). The results show that the success rate of recognition is 90%. It is proved that it is feasible to use the time-frequency characteristic of fault waveform to identify the fault type of distribution line.
【作者单位】： 上海交通大学电子信息与电气工程学院;云南电力科学研究院;国网河南省电力公司电力科学研究院;
【基金】：国家自然科学基金(51307109) 国家电网公司科技项目(SGTYHT/14-JS-188)资助
【分类号】：TM75
【参考文献】
相关期刊论文 前7条
1 谢潇磊;刘亚东;孙鹏;罗林根;盛戈v，
本文编号：2376034
本文链接：https://www.wllwen.com/kejilunwen/dianlidianqilunwen/2376034.html
