摘要
为预测电网谐波,改进供电品质,本文根据电网谐波特性,建立了BP神经网络算法,对电网谐波样本点进行了误差训练,得到了谐波幅值变化特性。通过与标准值比较,表明:采用BP神经网络算法对谐波进行拟合能够达到较高的准确度。
In order to predict the grid harmonic,and providing theoretical basis for improving power supply quality. According to the characteristics of the harmonic wave in the power grid, this paper establishes the BP neural network algorithm,and has carried on the training error to sample points of the harmonic wave in the power grid,harmonic amplitude variation characteristics are obtained finally. By comparison with the standard values,the BP neural network algorithm for harmonic fitting can reach high accuracy.
出处
《舰船科学技术》
北大核心
2016年第14期55-57,共3页
Ship Science and Technology
基金
河南省教育厅科学技术研究重点项目(13B520253)
河南省教育厅科学技术研究重点项目(13B890250)
