基于分层聚合的电力系统不良数据自动辨识  

Automatic identification of bad data in power system based on hierarchical aggregation

作　　者：翟佳 封冠华 张翃帆 ZHAI Jia;FENG Guanhua;ZHANG Hongfan(State Grid Beijing Customer Service Center,Beijing 100062,China)

机构地区：[1]国网北京客服中心,北京100062

出　　处：《电子设计工程》2023年第16期187-190,195,共5页Electronic Design Engineering

摘　　要：为了准确分辨电力系统中的不良数据,从而保证电网安全运行,提出了基于分层聚合的电力系统不良数据自动辨识方法。根据电力系统主电路连接形式,计算分层标准与聚合指标的实际数值,完成基于分层聚合的电力系统规划。在此基础上,确定电力数据的不良程度,通过数据加权残差的处理方式,得到具体的辨识标度值,利用该值自动辨识电力系统不良数据。对比实验结果显示,与前推回代追踪辨识方法相比,基于分层聚合的自动辨识方法的辨识准确率较高,能够保证电网安全运行,符合实际应用需求。In order to accurately distinguish the bad data in power system and ensure the safe operation of power grid,an automatic identification method of bad data in power system based on hierarchical aggregation is proposed.According to the main circuit connection form of power system,the actual values of hierarchical standard and aggregation index are calculated,and the power system planning based on hierarchical aggregation is completed.On this basis,the bad degree of power data is determined,and the specific identification scale value is obtained through the processing method of data weighted residual,which is used to automatically identify the bad data of power system.The comparative experimental results show that compared with the forward push back tracking identification method,the automatic identification method based on hierarchical aggregation has higher identification accuracy,can ensure the safe operation of power grid and meet the needs of practical application.

关 键 词：分层聚合 不良数据 自动辨识 加权残差 辨识标度值 

分 类 号：TN929[电子电信—通信与信息系统]