论文部分内容阅读
研究人工神经网络用于波形识别是一个新的研究领域。将神经网络这一新方法引入到电网谐波在线监测中,并就它在电网谐波在线监测应用中的一系列关键问题进行了较为系统的分析。提出了一种新型电网谐波在线监测系统模型,通过带通滤波器,采样存储器以及神经网络的在线监测电路,取得谐波信号,最终达到谐波补偿的目的。其在线监测部分主要利用模拟并行测量的基本原理,构造了一个用于电网谐波在线监测的特殊多层前向神经网络。并对其进行了仿真,仿真结果表明所提出的基于人工神经网络的电网谐波在线监测的方法是可行的和有效的。
Research on artificial neural network for waveform recognition is a new field of research. The new method of neural network is introduced into online monitoring of harmonic in power grid and a series of key problems in the online monitoring of power grid harmonic are systematically analyzed. A new model of on-line monitoring system for harmonic in power grid is proposed. By means of band-pass filter, sampling memory and online monitoring circuit of neural network, a harmonic signal is obtained and the harmonic compensation is finally achieved. Its on-line monitoring mainly uses the basic principle of analog parallel measurement to construct a special multi-layer neural network for on-line monitoring of power grid harmonics. The simulation results show that the proposed method based on artificial neural network for on-line monitoring of power network harmonics is feasible and effective.