摘要
针对特高压直流输电系统中的换流器区在故障发生后难以快速确定故障类型与位置,提出一种融合空洞卷积的MobileNetV2特高压换流器区故障识别方法 .首先通过特高压直流输电系统的PSCAD/EMTDC仿真模型采集换流器区故障电流信号作为故障识别的依据;然后利用小波分解对原始的故障信号进行处理,选取与原始信号相关性最大的分解信号作为网络模型的数据集,去除原始信号中的冗余量;最后将空洞卷积、Ghost模块和ULSAM注意力机制引入MobileNetV2网络中,对网络中倒残差结构的进行改进得到故障识别模型.实验结果表明,融合空洞卷积的MobileNetV2模型故障识别准确率达到98.75%,比其他网络模型有更好的故障识别效果.
In view of the difficulty in quickly determining the fault type and location in the converter area of the UHVDC transmission system after the fault occurs,a fault identification method for the MobileNetV2 UHV converter area with void convolution was proposed.Firstly,the PSCAD/EMTDC simulation model of the UHVDC transmission system was used to collect the fault current signal in the converter area as the basis for fault identification.Subsequently,wavelet decomposition is applied to the original fault signal,and the component exhibiting the highest correlation with the original signal is selected as the dataset for the network model,effectively reducing redundancy in the original signal.Finally,the dilated convolution,Ghost module and ULSAM attention mechanism are introduced into the MobileNetV2 network,and the inverted residual structure in the network is improved to obtain the fault identification model.Experimental results show that the fault identification accuracy of the proposed MobileNetV2 model fused with dilated convolution reaches 98.75%.Compared with other network models,the proposed model has better fault identification effect.
作者
向鑫
吴浩
李小鹏
宋弘
吴川兰
XIANG Xin;WU Hao;LI Xiaopeng;SONG Hong;WU Chuanlan(School of Automation and Information Engineering,Sichuan University of Science&Engineering,Yibin,Sichuan 644000,China;Intelligent Perception and Control Key Laboratory of Sichuan Province,Yibin,Sichuan 644000,China;State Grid Sichuan Electric Power Research Institute,Chengdu,Sichuan 610041,China;College of Electronic Information and Automation,Aba Teachers College,Aba,Sichuan 623002,China)
出处
《宜宾学院学报》
2025年第12期65-72,共8页
Journal of Yibin University
基金
人工智能四川省重点实验室项目(2023RYY06)
四川轻化工大学人才引进项目(2021RC12)
太阳能技术集成及应用推广四川省高校重点实验室(SN240102)。
