新型神经模糊控制的并联有源电力滤波器  

A Novel Shunt Active Power Filter Based on Neural Network and Fuzzy Control

作　　者：惠晶[1] 韩立圣[1] 王存款[1] 

机构地区：[1]江南大学,信控学院,江苏无锡214122

出　　处：《电力电子技术》2010年第10期44-47,共4页Power Electronics

摘　　要：针对有源电力滤波器(APF)中谐波电流检测的快速实时响应问题,提出了一种新型三相并联型APF的神经模糊控制方案,利用遗传算法进行训练。建立了三相并联型APF的数学模型,设计了神经模糊控制器,在以DSP为核心控制器的基础上对其进行了实验研究。仿真和实验结果表明,该方法实时快速性好、动态特性好,验证了方案的正确性和可行性。Against the problem of APF distortion in the current real-time response to the rapid detection,a novel neu-ral network and fuzzy control strategy is proposed,the genetic algorithm（GA）is used to train the network.The math-ematic model of APF is established and fuzzy neural network controller is designed.An experimental study about it is conducted based on DSP.Simulation and experimental results show that the method has a good real-time quality and good dynamic characteristics,which verify the feasibility and accuracy of the proposed method.

关 键 词：有源电力滤波器 神经模糊控制 谐波 遗传算法 

分 类 号：TN713[电子电信—电路与系统]