机构地区: 长沙铁道学院
出 处: 《长沙铁道学院学报》 1999年第3期54-58,111,共6页
摘 要: 基于自适应噪声对消技术及人工神经网络(ANN)理论,提出非线性负载高次谐波及基波无功电流动态检测的一种新方法.所设计的谐波检测系统在线训练ANN并具有二级滤波,即使非线性负载运行发生了变化,该系统仍能根据需要快速检测出以下多个参数:基波有功、无功电流,基波位移因数,特定的几次(如3、5、7次等)谐波电流及其余的谐波总电流等.仿真结果证实,该系统所检测出的各项参数的波形畸变和相移非常小.可应用于电力系统谐波补偿. Based on self-adaptive noise countervaailing m ethod and artificialneuralnet- w ork (ANN) theory, this paperproposes a new approach to dynam ic detect- ing harm onics and fundam ental reactive current of nonlinear loads. As the ANN is trained on line, the detecting system , w ith tw o-levelfilter, can de- tect the follow ing param eters: fundam entalactive and reactive current, fun- dam entalphase diaplacem ent factror, assigned orders (e.g.3, 5, 7 orders) harm onics and thetotalharm onics ofthe rest, even iftheoperation ofnonlin- ear olads changes abruptly. The sim ulation proves that the distortion and phase diaplacem entofthe ANN outputw aveform s arevery tiny. Thissystem can be applied in pow ersystem real-tim em onitoring and in harm oniccom pen- sation ofactive pow erfilter, especially ofhybrid active pow er filter.
领 域: [电气工程]