摘要
为实现气体绝缘开关设备GIS(gas insulated swithgear)运行状态的智能监测和有效辨识,提出一种基于格拉姆角场增强和残差神经网络的GIS运行状态辨识方法。该方法通过格拉姆角场对GIS设备的原始振动时域信号进行高阶表征,将其投影为二维图谱,再利用残差神经网络实现GIS设备的状态辨识。搭建了包含3种GIS设备典型机械缺陷的试验模拟测试系统,试验结果表明:所提方法不仅能够有效表征原始信号的状态特征,且辨识精度较常规方法有显著提升,提升约5%,证明了所提方法的有效性。
In order to realize intelligent monitoring and effective identification of operation status of gas insulated switchgear(GIS),in this paper a kind of operation status identification method of GIS based on enhanced Gramian an⁃gular fields and residual neural network are proposed.The original vibration time⁃domain signals of GIS equipment are characterized by high⁃order through the Gramian angular fields,and it is projected into a two⁃dimensional pic⁃ture,and then the residual neural network is used to achieve the status identification of GIS.A test simulation testing system including three typical mechanical defects of GIS is set up.The test results show that the proposed method can not only effectively characterize the status characteristics of the original signal,but also the identification accuracy is significantly improved by approximately 5%compared with the conventional method,which proves the effectiveness of the proposed method.
作者
王劭鹤
赵琳
杨勇
金涌涛
王枭
陈孝信
张阳
WANG Shaohe;ZHAO Lin;YANG Yong;JIN Yongtao;WANG Xiao;CHEN Xiaoxin;ZHANG Yang(Electric Science Research Institute of Zhejiang Provincial Electric Company,Hangzhou 310014,China;Shanghai Ruishen Electronic Technology Co.,Ltd.,Shanghai 201108,China)
出处
《高压电器》
北大核心
2025年第8期39-46,共8页
High Voltage Apparatus
基金
国网浙江省电力有限公司科技项目(5211DS20008K)。
关键词
振动信号
格拉姆角场
气体绝缘开关
残差神经网络
状态辨识
vibration signals
Gramian angular fields
gas insulated switchgear
residual neural network
status identification
