基于弹性网络的空心线圈电流互感器误差预测  

Error Prediction of Air-core-coil Current Transformer Based on Elastic Network

作　　者：李振华[1,2] 陈兴新 蒋伟辉 卢梦瑶 李振兴[1] LI Zhenhua;CHEN Xingxin;JIANG Weihui;LU Mengyao;LI Zhenxing(Hubei Key Laboratory for Operation and Control of Cascade Hydropower Stations,Three Gorges University,Hubei Yichang 443002,China;School of Electrical and new Energy,Three Gorges University,Hubei Yichang 443002,China)

机构地区：[1]三峡大学梯级水电站运行与控制湖北省重点实验室,湖北宜昌443002 [2]三峡大学电气与新能源学院,湖北宜昌443002

出　　处：《高压电器》2022年第9期134-141,共8页High Voltage Apparatus

基　　金：国家自然科学基金资助项目(51877122)。

摘　　要：针对空心线圈电流互感器长期可靠性和稳定性不高的问题,文中首先利用皮尔逊相关系数和基于弹性网络因子筛选的方法,得到影响空心线圈电流互感器误差的主导因素是环境参量温度和电气参量负荷;再利用机器学习的线性回归模型弹性网络对空心线圈电流互感器进行误差预测,同时为避免陷入局部最小值和过拟合的问题,使用了交叉验证的方法。在此基础上,将所提模型的预测结果与传统SVM和KNN算法的预测结果进行了对比分析,实例结果表明弹性网络对误差的预测有更高的准确性、稳定性和解释性。In view of issue as low long term reliability and stability of air core coil current transformer,the Pearson correlation coefficient and the method based on elastic network factor screening is used firstly in this paper.It is con-cluded that the main factors affecting the error of air core coil current transformer are environmental parameter tem-perature and electrical parameter load.Then,the elastic network of linear regression model of machine learning is used to predict the error of air core coil current transformer.At the same time,the cross validation method is used in order to avoid falling into the local minimum and overfitting.On this basis,the prediction results of the proposed mod-el are compared with those of the traditional SVM and KNN algorithm.The example results show that the elastic net-work has higher accuracy,stability and explanation to the error prediction.

关 键 词：弹性网络 误差预测 变量筛选 空心线圈电流互感器 

分 类 号：TM452[电气工程—电器]