中压配电树线矛盾林火故障的导线痕迹图像特征及其识别方法研究  

Research on image characteristics and identification methods of conductor trace caused by forest fires of tree-line conflicts in MV distribution overhead lines

作　　者：苏俊 李天友 陈石川 郭伯森 黄超艺 SU Jun;LI Tianyou;CHEN Shichuan;GUO Bosen;HUANG Chaoyi(College of Electrical Engineering and Automation,Xiamen University of Technology,Xiamen 361024,China;Xiamen Key Laboratory of Frontier Electric Power Equipment and Intelligent Control,Xiamen 361024,China;State Grid Fujian Electric Power Company,Fuzhou 350003,China;State Grid Quanzhou Power Supply Company,Quanzhou 362101,China)

机构地区：[1]厦门理工学院电气工程与自动化学院,福建厦门361024 [2]厦门市高端电力装备及智能控制重点实验室,福建厦门361024 [3]国网福建省电力有限公司,福建福州350001 [4]国网福建省电力有限公司泉州供电公司,福建泉州362101

出　　处：《供用电》2024年第11期20-30,共11页Distribution & Utilization

基　　金：国家电网有限公司总部科技项目(5500-202221138A-1-1-ZN)。

摘　　要：中压配电线路常跨越植被茂密的林区,这些区域极易发生山火,严重影响电网的安全稳定运行。为给中压线路山火事故的鉴别提供参考依据,提出一种基于中压线路放电及火焰作用后导线痕迹图像特征及其识别方法的研究。首先,通过树线矛盾真型试验平台模拟中压架空线路对树放电以及受火焰作用下的导线痕迹,基于场景一(中压线路火焰作用)和场景二(中压线路树线放电)的中压架空裸导线提取了火焰作用和树线放电对导线造成损伤的痕迹图像特征,并以此建立树线矛盾故障特征图谱库。利用多维特征分析技术,对比研究了放电及火焰作用后两者导线痕迹之间的相似性和差异性,并通过相关性分析确定了影响导线痕迹的关键图像特征参量。最后,利用支持向量机(support vector machine,SVM)算法提出基于关键图像特征参量的导线燃烧及放电痕迹图像识别技术,通过导线痕迹辨别树线矛盾故障事故的起因。该研究结果可为中压线路在火场环境中燃烧痕迹的分析及起火原因的判别提供参考依据。Medium voltage distribution overhead lines frequently traverse dense ly vegetated forest areas and regions with intensive human agricultural activity,making them highly susceptible to wildfires that can severely impact the safety and stability of electrical grid operations.To provide a reference for identifying incidents of wildfires affecting medium voltage overhead lines,this paper development a research on image characteristics and identification methods of conductor trace based on discharges of mediumvoltage overhead lines and flames influence.Initially,using the real-type experimental platform for tree-line conflicts,simulations were conducted to observe the discharge between medium-voltage overhead lines and trees,as well as the traces left on conductors under the influence of flames.Based on scenario one(flame influence on medium-voltage lines)and scenario two(tree-line discharge of medium-voltage lines),the image features of damage traces on medium-voltage overhead bare conductors were extracted.These features were then used to establish a fault characteristic atlas for tree-line conflicts.Utilizing multidimensional feature analysis techniques,the study compared the similarities and differences between wire traces after discharge and flame influence,and through correlation analysis,identified key image feature parameters that influence wire trace characteristics.Finally,employing the support vector machine(SVM)algorithm,a technique for image recognition of wire burning and discharge traces based on key image feature parameters was proposed,which facilitates the determination of the causes behind tree-line conflict accidents.The results of this research can provide a basis for analyzing burn traces and determining the causes of fires in medium voltage overhead lines in fire-prone environments.

关 键 词：中压线路 火焰作用 导线痕迹 图像特征 识别方法 

分 类 号：TM71[电气工程—电力系统及自动化]