Skip to Main Content
A new complete RBF neural network SVPWM controller scheme for a grid-connected inverter is presented in this paper, RBF network architecture and the parameter is determined by using the clustering method, the RBF neural network which is simplifies and realize voltage space vector modulate to control the inverter. It was used in three-phase photovoltaic grid-connected inverse system which adopts control of predictive current. A computer simulation program is developed using MATLAB. The simulation results show that the method features simplicity, high efficiency and excellent control effect.
Date of Conference: 6-7 April 2009