基于半监督机器学习的复杂电网连锁故障诊断方法  

The Diagnosis of Complex Grid Chain Faults Based on Semi-supervised Machine Learning

作　　者：左娟娟 朱红杰 杨继党 张华生 肖航 ZUO Juan-juan;ZHU Hong-jie;YANG Ji-dang;ZHANG Hua-sheng;XIAO Hang(Baoshan Power Supply Bureau of Yunnan Power Grid Co.,Ltd.,Baoshan 678000 China)

机构地区：[1]云南电网有限责任公司保山供电局,云南保山678000

出　　处：《自动化技术与应用》2024年第12期47-50,92,共5页Techniques of Automation and Applications

摘　　要：复杂电网连锁故障诊断性能过差会降低电力系统运行效率,为了及时监测复杂电网连锁故障,提升电力系统运行效率,提出基于半监督机器学习的复杂电网连锁故障诊断方法。根据电网负荷状态和电力信息物理系统建立复杂电网连锁故障模型,利用交叉小波和主元分析提取模型内故障特征。将故障特征与隶属于半监督机器学习的LCop K-means算法结合,根据算法输出的分类结果实现复杂电网连锁故障诊断。实验结果表明,所提方法诊断延迟较小,且故障区域的定位能力较强。The poor performance of complex power grid cascading fault diagnosis will reduce the efficiency of power system operation.In order to monitor complex power grid cascading faults in time and improve the efficiency of power system operation,a complex power grid cascading fault diagnosis method based on semi supervised machine learning is proposed.According to the load state of the power grid and the physical system of power information,a complex power grid cascading fault model is established,and the fault characteristics in the model are extracted by using Cross Wavelet and principal component analysis.The fault characteristics are combined with lcop k-means algorithm which belongs to semi supervised machine learning,and the cascading fault diagnosis of complex power grid is realized according to the classification results output by the algorithm.The experimental results show that the diagnosis delay of the proposed method is small,and the fault location ability is strong.

关 键 词：复杂电网 故障诊断 模型 交叉小波 主元分析 半监督机器学习 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] TM734[自动化与计算机技术—控制科学与工程]