基于INGO-SVM的输电铁塔地脚螺栓螺母缺失无损检测方法  

Non-destructive Detection Method for Nuts Missing Defect on Anchor Bolts of Transmission Tower Based on INGO-SVM

作　　者：刘阳 张璐 吴德强 周青 张川 王彦海[2] LIU Yang;ZHANG Lu;WU Deqiang;ZHOU Qing;ZHANG Chuan;WANG Yanhai(Yichang Power Supply Company of State Grid Hubei Electric Power Co.,Hubei Yichang 443000,China;College of Electrical Engineering&New Energy,China Three Gorges University,Hubei Yichang 443000,China)

机构地区：[1]国网湖北省电力有限公司宜昌供电公司,湖北宜昌443000 [2]三峡大学电气与新能源学院,湖北宜昌443000

出　　处：《高压电器》2025年第2期130-140,共11页High Voltage Apparatus

基　　金：国家自然科学基金资助项目(U22A20600,52079070)。

摘　　要：为了无损检测埋置于混凝土中的输电铁塔地脚螺栓有无螺母缺失缺陷,保障输电线路安全运行,文中提出了一种基于改进北方苍鹰优化算法(improve northern goshawk optimization,INGO)优化的支持向量机(support vector machine,SVM)分类检测方法(INGO-SVM):首先,通过Cubic混沌映射与小孔成像反向学习策略增加北方苍鹰优化算法(northern goshawk optimization,NGO)种群的多样性,并在优化初始解的同时增加种群的搜索区域,使算法尽可能的找到潜在的最优解并分析优化效果;其次,将INGO应用于SVM的核心参数寻优,得到分类模型;最后,将螺杆直径、保护层厚度、垫板厚度以及电磁无损检测得到的磁场强度作为输入量,输出地脚螺栓螺杆上螺母个数,判断缺陷类型;实验结果表明,相较于SVM,提出的INGO-SVM模型在输电铁塔地脚螺栓螺母缺失分类中的均方根误差、平均相对误差以及平均绝对误差分别降低了31.7%、60.7%、68.9%,验证了该方法解决地脚螺栓螺母缺失无损检测分类问题的有效性。For non-destructively detecting the missing and defect of nuts on the anchor bolts of the transmission tower buried in the concrete and ensuring safe operation of the transmission line,a kind of support vector machine(SVM)classification detection method based on an improved northern goshawk optimization(INGO)is proposed in this paper.Firstly,Cubic chaotic mapping and pinhole imaging reverse learning strategy are used to increase the diversity of the northern goshawk optimization(NGO)population and,at the same time of optimizing the initial solution,expand the search area of the population,increase the search area for the population and make the algorithm find potential optimal solutions as much as possible and analyze optimization results.Secondly,INGO is applied to optimize the core parameters of SVM to obtain the classification model.Finally,the screw diameter,protective coating thickness,gasket thickness,and the magnetic field strength detected by electromagnetic non-destructive testing are used as the input,the number of nuts on the anchor bolt of the screw is output to determine the type of defect.The experimental results show that compared with SVM,the root mean square error,average relative error and average absolute error of the proposed INGO-SVM in the missing anchor bolt and nut classification of transmission tower are reduced by 31.7%,60.7%and 68.9%,respectively.The effectiveness of the method in solving the non-destructive detection and classification problem of missing nuts in foundation bolts is verified.

关 键 词：输电铁塔地脚螺栓 螺母缺失缺陷 改进北方苍鹰优化算法 支持向量机 电磁无损检测 

分 类 号：TU317[建筑科学—结构工程] TM75[电气工程—电力系统及自动化]