基于ANN的补偿电流互感器饱和对传变电流影响的算法

摘要

A new algorithm based on artificial neural network (ANN) is developed to eliminate and compensate the current transforming error caused by current transformer saturation. The essential idea is that, when the current transformer is saturated, the current of primary side can be calculated through the current of secondary side using the non-linear mapping ability of ANN. Thus the influence by the saturation of current transformer will be eliminated. Comparison between the proposed algorithm based on ANN and the algorithm based on differential equation is also given in this paper. The result shows that ANN based algorithm is better than differential equation based algorithm.%提出一种基于人工神经网络(ANN)的消除和补偿电流互感器饱和对传变一次电流影响的算法。算法的基本思想为：当电流互感器饱和时，利用人工神经网络极强的非线性映射能力，由电流互感器的二次电流准确计算出一次电流，从而有效地消除和补偿了电流互感器饱和对传变一次电流的影响。与传统的基于微分方程的补偿算法进行了比较，结果表明基于ANN的算法能够获得比采用传统的微分方程算法更好的结果。