多源信号特征融合的电能质量扰动识别  

Power quality disturbance identification based on multi-source signal feature fusion

作　　者：陈思源 程志友 杨猛[1] 胡乐乐 CHEN Siyuan;CHENG Zhiyou;YANG Meng;HU Lele(School of Electronic Information Engineering,Anhui University,Hefei 230601,China;School of Internet,Anhui University,Hefei 230039,China;Ministry of Education Engineering Research Center of Power Quality,Anhui University,Hefei 230601,China)

机构地区：[1]安徽大学电子信息工程学院,安徽合肥230601 [2]安徽大学互联网学院,安徽合肥230039 [3]安徽大学教育部电能质量工程研究中心,安徽合肥230601

出　　处：《安徽大学学报（自然科学版）》2024年第4期62-66,共5页Journal of Anhui University(Natural Science Edition)

基　　金：国家自然科学基金资助项目(6227020935);安徽省科技重大专项(18030901018)。

摘　　要：为了解决风能、太阳能等可再生能源输出的不稳定性和间歇性给电能质量带来的问题,提出多源信号特征融合的电能质量扰动识别方法.该方法引入电流信息增强扰动特征,为解决电能质量扰动识别提供了新的视角.算例分析结果表明:相对于其他2种方法,该文方法的4个评价指标(准确率、精确率、召回率和F1分数)均最高.因此,该文方法具有优越性.To address the issues of instability and intermittency in the output of renewable energy sources such as wind and solar power affecting power quality,a method for power quality disturbance identification based on multi source signal feature fusion was proposed,This method introduced current information to enhance disturbance features,providing a new perspective for solving power quality disturbance identification.Case analysis results indicated that compared with two other methods,the proposed method achieved the highest scores in four evaluation metrics(accuracy,precision,recall and F1 score).Therefore,the proposed method demonstrated superiority.

关 键 词：电能质量扰动 残差网络 多源信号特征融合 相对位置矩阵 有效通道注意力 

分 类 号：TM712[电气工程—电力系统及自动化]