光伏并网发电系统最大功率点跟踪算法研究.pdfVIP

20 9 Vol. 20 No. 9 2010 9 COM PUT ER TECHNOLOGY AND DEVELOPMENT Sep. 2010 申 楠, 段中兴 (西安建筑科技大学信息与控制工程学院, 陕西 西安 710055) : , , , , , : , , : ; ; ; : T P183 : A : 1673- 629X( 2010) 09- 0197- 04 Research on Grid- connected PV System Maximum Power Point Tracking Algorithm SHEN Nan, DUAN Z ongxing ( Sc ool of Information and Control Engineering, Xi an University of Arc itecture and Tec nology, Xi an 710055, C ina) Abstract:Aiming at t e disadvantage of low convergence speed and local minimum of neural networks algorit m used in t e solar cell max imum pow er point tracking ( MPPT ) , an effective neural network algorit m based on genetic algorit m ( GA) is proposed to realize t e M PPT control in t is paper. Adopting t e global searc ing c aracteristic of genetic algorit m and genetic operations suc as selection, crossing and mutation, an optim ization of w eig ts in neural networks are ac ieved, and t e negative effects of netw ork s initial w eig ts on control effect are solved. T e simulation and experiments s ow t at t e algorit m can improve t e convergence speed and nonlinear ap proximation ability in t e MPPT control, and quickly and accurately track MPPT in t e suns ine intensity, ambient temperature c ang ing environment. Key words:p otovoltaic grid- connected; MPPT; neural networks; genetic algorit m [ 1, 2] , , , [ 3] , , Cuk , , , , , ; [ 4]