基于LK-means算法的绝缘子冰闪放电的声发射信号检测技术研究  

Research on Ice Flash Discharge Detection of Insulators Based on LK-means Clustering Algorithm

作　　者：梁冰 朱海涛 余洋 黄斯凯 扈海泽 LIANG Bing;ZHU Haitao;YU Yang;HUANG Sikai;HU Haize(State Grid Hunan Extra High Voltage Transmission Company,Hengyang 421000,China;School of Computer Science and Engineering,Hunan University of Science and Technology,Xiangtan 411201,China)

机构地区：[1]国网湖南省电力有限公司超高压输电公司,湖南衡阳421000 [2]湖南科技大学计算机科学与工程学院,湖南湘潭411201

出　　处：《电力学报》2023年第5期374-381,共8页Journal of Electric Power

基　　金：国家自然科学基金(61866013)。

摘　　要：绝缘子冰闪放电是一个长期积累的过程,会对电力系统的稳定运行造成巨大威胁,但很难对其进行准确的检测。为了降低绝缘子冰闪的影响,提出了一种基于优化聚类算法和深度学习模型的自适应冰闪预警新方法。首先,研究了绝缘子的冰闪放电过程,分析了不同阶段绝缘闪络时所产生的声发射信号的特性;其次,从无人机采集的声发射信号中提取MFCC参数,通过LK-means聚类算法对特征量进行分类;最后,采用卷积神经网络对各个放电阶段的声信号特征进行提取,从而自主识别采集信号所处阶段。为了验证所提方法的有效性,在MATLAB平台上进行了试验仿真分析。试验结果表明,所提方法对绝缘子冰闪放电过程的预警故障识别准确率均达到96%以上,为输电线路观冰及调整负荷提供参考依据。所提方法能准确地检测绝缘子冰闪故障,降低了绝缘子冰闪事故发生的概率,提高了电力系统供电可靠性。Ice flashover discharge of insulator is a process of long-term accumulation,which is beneficial to the stable operation of power system,but it is difficult to detect it accurately.In order to reduce the impact of insula-tor ice flashover,a new method of adaptive ice flashover warning based on optimized clustering algorithm and deep learning model is proposed.Firstly,the ice flashover discharge process of insulators is studied,and the characteristics of acoustic emission signals generated by insulation flashover at different stages are analyzed.Sec-ondly,the MFCC parameters are extracted from the acoustic emission signals collected by the UAV,and the feature quantities are classified by the LK-means clustering algorithm.Finally,the convolutional neural network is used to extract the acoustic signal characteristics of each discharge stage,so as to independently identify the stage of the collected signal.In order to verify the effectiveness of the proposed method,the experimental simu-lation is designed on the MATLAB platform.The experimental results show that the proposed method achieves an accuracy of more than 96%for the identification of insulator ice flash discharge over warning faults,which provides a reference basis for ice observation and load adjustment of transmission lines.Therefore,the proposed method can accurately detect insulator ice flash faults,reduce the probability of insulator ice flash acci-dents,reduce the accident rate of power system,and thus improve the power supply reliability of power system.

关 键 词：绝缘子 覆冰闪络 LK-means MFCC 卷积神经网络 

分 类 号：TM216[一般工业技术—材料科学与工程]