基于BP神经网络的关口电能计量装置测量误差预测及校正

Error prediction and correction of gate energy measurement based on BP neural network

关口电能计量装置造成的误差通常会给电网公司带来巨大的经济损失,因此提高关口电能计量装置的准确度,具有十分重要的应用价值.通过对关口电能计量装置的历史数据进行分析,采用BP(back propagation)神经网络算法进行误差预测,筛选出最适合关口电能计量数据的优化模型,并且校正计量异常值,从而减小电能计量装置产生的误差,提高电能计量的准确性.实验表明,误差预测及校正模型能准确预测关口电能计量装置误差,修正异常值.

Improving the accuracy of gate energy measurement is very important and worthwhile. In this paper,the historical data of the gate energy measurement are analyzed,and the BP(Back Propagation) neural network algorithm is used to predict the error. The optimal model of the metrological data is selected to correct the abnormal value,thus reducing the impact of the gate energy measurement error,and improving the accuracy of energy measurement. Experiments show that this error prediction and correction model can accurately predict the error of the gate energy measurement and correct the abnormal value.