基于Teager能量算子和1D-CNN的HVDC输电线路故障识别方法  

Fault Identification Method of HVDC Transmission Line Based on Teager Energy Operator and 1D-CNN

作　　者：王桥梅 吴浩 杨杰 李栋 刘益岑[3] WANG Qiaomei;WU Hao;YANG Jie;LI Dong;LIU Yicen(College of Automation and Information Engineering,Sichuan University of Science&Engineering,Zigong 643000,China;Artificial Intelligence Key Laboratory of Sichuan Province,Zigong 643000,China;State Grid Sichuan Electric Power Research Institute,Chengdu 610000,China)

机构地区：[1]四川轻化工大学自动化与信息工程学院,四川自贡643000 [2]人工智能四川省重点实验室,四川自贡643000 [3]国网四川电力科学研究院,四川成都610000

出　　处：《智慧电力》2021年第5期93-100,共8页Smart Power

基　　金：国家自然科学基金资助项目(11705122);四川省科技厅项目(2017JY0338,2019YJ0477,2018GZDZX0043);四川轻化工大学研究生创新基金(y2019012)。

摘　　要：针对HVDC输电线路故障识别率低,远端高阻故障识别困难等问题,提出基于Teager能量算子和1DCNN的HVDC输电线路故障识别方法。该方法利用保护安装处测得的线模分量Teager能量算子和输电线路两侧正负极电流突变量能量比值组成特征向量,利用1D-CNN对特征向量集进行训练和测试,同时实现区内外故障判和故障极选择。仿真实验表明该方法能在不同故障距离和不同过渡电阻情况下有效实现区内外故障识别和故障极选择,采样率能满足现有实际工程需要,具有较强的耐受过渡电阻能力。Aiming at the low fault recognition rate of the existing HVDC transmission line protection and the difficulty of remote highimpedance fault recognition,a fault recognition method of HVDC transmission line based on Teager energy operator and 1 D-CNN is proposed.This method uses the line mode component Teager energy operator measured at the protection installation and the energy ratio of the positive and negative current mutations on both sides of the transmission line to form a feature vector,and 1 D-CNN is used to train and test the feature vector set,the fault judgment and fault pole selection inside and outside the region are realized at the same time.A large number of simulation experiments show that the method fault identification and fault pole selection under different fault distances and different transition resistances can be realized effectively.The sampling rate can satisfy the real project requirements.The method has strong ability to withstand transition resistance.

关 键 词：HVDC 输电线路 TEAGER能量算子 1D-CNN 故障识别 

分 类 号：TM773[电气工程—电力系统及自动化]