三相四线制APF电流预测控制策略仿真  

Research on Three-phase Four-wire APF Based on Predictive Current Control Trategy

作　　者：陈良[1] 黄传金[1] 曹文思 时伟[1] 

机构地区：[1]中州大学电气教研室 [2]华北水电水利学院电气教研室

出　　处：《科学技术与工程》2012年第22期5469-5473,共5页Science Technology and Engineering

基　　金：郑州市普通科技攻关项目(121PPTG364-3)资助

摘　　要：针对三相四线制有源电力滤波器(APF)固有延时造成APF补偿性能下降问题,提出了一种基于神经网络的电流预测控制方法。研究了基于神经网络模型预测控制的基本原理。建立了三相四线制APF电流预测控制结构模型。在此基础上设计了神经网络预测控制器。实现了对三相四线APF电流的预测控制。仿真实验结果证明,基于神经网络的电流预测控制具有较好的跟踪能力,克服了网侧电流存有毛刺的现象。提高了APF补偿能力。Focusing on the problem of reduction of the APF compensating ability caused by the system inherit time delay of the Active Power Filter, a control strategy is givh based on NNPC which is studied deeply in basic principle, designs the NNPC model on which the predictive current control of the three-phase four-wire APF is achieved. The simulation results prove that the NNPC strategy has the benefit of good tracing capacity and eliminates the current glitch on the network side.

关 键 词：三相四线制 源电力滤波器 预测控制 神经网络 

分 类 号：TM762.3[电气工程—电力系统及自动化]