摘要
提出了一种基于人工神经网络 ( ANN)的电力系统谐波测量新方法。该方法应用一个多层前馈神经网络 ( MLFNN) ,对当前采样时刻和上一采样时刻的三相电流采样值进行分析计算 ,得出三相电流的谐波分量。阐述了该神经网络的构造和用于网络训练的学习算法。将该网络应用于整流电路的谐波测量 ,进行了仿真研究。仿真结果表明该方法能够实时而准确地检测出谐波分量。通过与基于瞬时无功功率理论的谐波测量方法比较 。
A new approach based on artificial neural network (ANN) for measuring power harmonics is proposed.It analyses and calculates the three-phase current of current sample time and of the last by a multi-layer feedforward neural network(MLFNN),then the harmonic components of the three-phase current are got.The structure of this neural network and the learning algorithm for training it are presented.The simulation studies are carried out by using this neural network into harmonic measuring of rectifier.The simulation results show that the harmonic components can be detected at real time with high precision.It is validated further that this approach is more precise with shorter delay compared with that based on instantaneous reactive power theory.
出处
《电力系统及其自动化学报》
CSCD
2004年第2期40-43,共4页
Proceedings of the CSU-EPSA
