摘要
在电能表异常识别中能够区分的异常数据种类少,异常数据识别时误差较大,因此,设计一种新的间接接入式直流电能表异常计量数据识别算法。对电能表计量数据进行归一化处理,引入k-means聚类,计算欧氏距离完成间接接入式直流电能表全部计量数据的聚类,优化小波变换得到离散小波,处理计量数据残差序列,提取异常计量数据特征进行异常计量数据识别。实例测试结果表明,该算法的最大误差为0.103,且波动较小,输出正常计量数据与异常计量数据,能够达到优化间接接入式直流电能表异常计量数据识别效果的目的。
In the abnormal identification of electric energy meter,there are few types of abnormal data that can be distinguished,and the error in abnormal data identification is large.Therefore,a new indirect access abnormal metering data identification algorithm of DC electric energy meter is designed.This paper normalizes the metering data of electric energy meter,introduce k-means clustering,calculate Euclidean distance,complete the clustering of all metering data of indirect access DC electric energy meter,optimize wavelet transform,obtain discrete wavelet,process the residual sequence of metering data,extract abnormal data characteristics,and identify abnormal metering data.The example test results show that the maximum error of the algorithm is 0.103,and the fluctuation is small.The output of normal measurement data and abnormal measurement data can achieve the purpose of optimizing the identification effect of abnormal measurement data of indirect access DC electric energy meter.
作者
崔胜胜
李汐
牟颖莹
李振
CUI Shengsheng;LI Xi;MOU Yingying;LI Zhen(Qinghai Marketing Service Center of State Grid Qinghai Electric Power Company,Xining 810000,China;Yuanqi Industrial Technology Co.,Ltd.,Qingdao 266000,China)
出处
《微型电脑应用》
2024年第7期130-133,共4页
Microcomputer Applications
