基于SHO优化SVM的输电线路雷电过电压识别  

Lightning Overvoltage Identification of Transmission Lines Based on SHO Optimized SVM

作　　者：马路金 李国龙[1] 邱世善 张志丽 MA Lu-jin;LI Guo-long;QIU Shi-shan;ZHANG Zhi-li(The Meteorological Agency of Huazhou City,Huazhou 525100,China)

机构地区：[1]广东省化州市气象局,广东化州525100

出　　处：《电气开关》2023年第3期52-55,59,共5页Electric Switchgear

摘　　要：为提高输电线路雷电过电压识别结果的准确性,以能量谱雷击波头、时域波形、时频谱等特征信息为输入量,采用斑点鬣狗算法对支持向量机进行优化,建立基于斑点鬣狗算法优化支持向量机的输电线路雷电过电压识别模型。仿真分析结果表明,SHO-SVM对非故障性过电压、感应雷过电压、反击故障过电压和绕击故障过电压识别的正确率分别为100%、100%、95%和95%,综合正确率高达97.5%,验证了模型的正确性和优越性。In order to improve the accuracy of transmission line lightning overvoltage identification results,taking the energy spectrum lightning head,time domain waveform,time spectrum and other characteristic information as input,the spotted hyena algorithm is used to optimize the support vector machine,and the transmission line lightning overvoltage identification model based on the spotted hyena algorithm optimized support vector machine is established.The simulation results show that the accuracy of SHO-SVM in identifying non fault overvoltage,induced lightning overvoltage,counterattack fault overvoltage and bypass fault overvoltage is 100%,100%,95%and 95%respectively,and the comprehensive accuracy is as high as 97.5%,which verifies the correctness and superiority of the model.

关 键 词：雷电过电压 输电线路 斑点鬣狗算法 支持向量机 识别 

分 类 号：TM86[电气工程—高电压与绝缘技术]