In order to ensure the accurate measurement of electric energy,this paper proposes an automatic compen-sation 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 coeffi-cient 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 com-pensation 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 fluc-tuation 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
Key words
transformer/error compensation/transfer entropy/SVR/LSTM/autoregressive integral moving average model
维普
万方数据