空间矢量和神经网络控制四桥臂APF电流  

Four-leg APF current based on space vector and neural network control

作　　者：武昌俊[1] 刘军[1] 王群京[2] 朱文武[1] 张莉[1] 

机构地区：[1]安徽机电职业技术学院,安徽芜湖241000 [2]合肥工业大学,安徽合肥230009

出　　处：《电源技术》2012年第4期543-545,549,共4页Chinese Journal of Power Sources

基　　金：安徽省高等学校省级自然科学研究项目(KJ2010B032)

摘　　要：提出了一种三相四桥臂有源滤波器(APF)的混合控制方法,即将四桥臂变流器控制分为三相主功率桥臂和第四桥臂控制两部分,前者采用矢量电流跟踪控制策略,同时,对第四桥臂负载零序电流的跟踪控制,采用神经网络控制器代替滞环比较器,不仅和滞环比较器一样简单可靠、实时性好,同时也避免了滞环控制造成的系统开关频率不固定,稳定性低等缺点。最后进行了仿真,结果表明效果良好。The hybrid control on three-phase four-leg APF was proposed, and namely the four-leg converter control was divided into three-phase main power leg control and the fourth leg control. The former adopted vector current tracking control and the zero-sequence tracking control of the four leg＇s loading adopted neural network control instead of hysteresis. Comparing with the hysteresis, the neural network control had the advantages of simple structure, good reliability and excellent real-time control, and avoided the disadvantages of the on-off frequency uncertainty and the poor stability caused by hysteresis. At last, the simulation result proves the hybrid control works well.

关 键 词：有源滤波器 四桥臂 矢量电流跟踪控制 神经网络控制 

分 类 号：TM712[电气工程—电力系统及自动化]