基于神经元网络的用于有源电力滤波器的电流检测  

AN ADAPTIVE DETECTING APPROACH OF CURRENTS BASED ON NEURAL NETWORK FOR ACTIVE POWER FILTER

作　　者：高大威[1] 孙孝瑞[1] 

机构地区：[1]华北电力大学电力工程系,北京100085

出　　处：《电网技术》2000年第1期72-75,共4页Power System Technology

摘　　要：文章将自适应噪声对消技术用于有源滤波器的电流检测，提出了一种基于神经元网络的自适应检测方法。该方法具有实时、准确、能自适应跟踪负载电流的变化等特点，并能在电力系统三相负载不对称情况下检测出负序和零序电流。理论分析和仿真结果证实了该方法的可行性。Using adaptive noise cancelling technology, an adaptive detecting approach of currents based on neural network for active power filter is proposed in this paper. Except the advantages such as real time, high accuracy, adaptability to load currents, etc., the proposed detecting approach also can be used to detect negative sequence and zero sequence currents under the condition of unbalanced load currents.The results of the theoretical analysis and computer simulation confirm the validity of the approach.

关 键 词：有源电力滤波器 神经元网络 电流检测 电力系统 

分 类 号：TM714.3[电气工程—电力系统及自动化] TN713.8[电子电信—电路与系统]