机构地区: 长沙铁道学院信息工程学院
出 处: 《铁道学报》 2000年第1期40-43,共4页
摘 要: 基于自适应噪声对消技术及人工神经网络(ANN)理论,提出了一种谐波电流动态检测方法。所设计的谐波检测系统采取在线学习、二级ANN滤波技术,能检测出所设定的n次(如3、5、7次等)谐波电流及剩余的谐波总电流,并能同时测出基波有功、无功电流和基波位移因数。仿真结果证实,该系统所测出的各项参数与实际值的相移和畸变非常小,且系统的结构简单,计算量小。该方法可应用于有源滤波或混合有源滤波的谐波及无功补偿。 Based on self adaptive noise countervailing method and artificial neural network (ANN) theory, this paper proposes a new approach to the dynamic detecting harmonics. As the ANN is trained on line, the detecting system with two level filter can detect assigned orders (e.g. 3,5,7 orders) harmonics, the total harmonics of the rest, the fundamental active and the reactive current and fundamental phase displacement factor. The simulation proves that the distortion and phase displacement of the ANN output waveforms are very tiny. The system can be applied in harmonic compensation of active power filter or hybrid active power filter and in reactive power compensation.