基于多种模型融合的电能计量互感器误差自动化补偿方法  

Automatic Compensation Method for Error of Electric Energy Measurement Transformers Based on Multiple Model Fusion

作　　者：杜瀚霖 DU Hanlin(Liuzhou Institute of Metrology and Measurement Technology,Liuzhou 545000,China)

机构地区：[1]柳州市计量技术测试研究所,柳州545000

出　　处：《自动化与仪表》2024年第9期110-113,117,共5页Automation & Instrumentation

摘　　要：为电能精准计量提供保障,提出基于多种模型融合的电能计量互感器误差自动化补偿方法。获取影响电能计量互感器误差的关键特征数据,作为改进LSTM与SVR输入,得到对应子模型的预测输出,计算子模型的权重系数,通过平均加权法融合模型的预测结果,确定电能计量互感器运行数据预测值,利用自回归积分滑动平均模型实现电能计量互感器误差的动态补偿。实验结果表明,特征数据选择可有效降低电能计量互感器比差和角差波动幅度;误差补偿后的二次电压曲线平滑,且更加接近理想的正弦波形。In order to ensure the accurate measurement of electric energy,this paper proposes an automatic compensation method for the error of electric energy metering transformer based on the fusion of multiple models.The key characteristic data selection that affects the error of electric energy metering transformer is obtained as the input of improved LSTM and SVR,and the predicted output of the corresponding sub-model is obtained,and the weight coefficient of the sub-model is calculated.The predicted value of the operation data of electric energy metering transformer is determined by fusing the prediction results of the model by the average weighting method,and the dynamic compensation of the error of electric energy metering transformer is realized by using the autoregressive integral moving average model.The experimental results show that the selection of characteristic data can effectively reduce the fluctuation range of ratio difference and angular difference of electric energy metering transformer.The secondary voltage curve after error compensation is smooth and closer to the ideal sine wave.

关 键 词：互感器 误差补偿 传递熵 SVR LSTM 自回归积分滑动平均模型 

分 类 号：TM45[电气工程—电器] TP29[自动化与计算机技术—检测技术与自动化装置]