Simulation of neuro-fuzzy controlled grid interactive inverter

In this study, a grid interactive inverter which is capable of importing electrical energy, generated from renewable energy sources such as wind, solar and fuel cells, to the grid line is designed and simulated with MATLAB/Simulink. A line frequency transformer and LCL filter is used at the output of the grid interactive inverter that is designed as voltage source inverter. Inverter is designed as current controlled to decrease the susceptibility to phase errors, and neuro-fuzzy controller is used as current controller. The membership functions and the rule base of the fuzzy logic controller, which control the inverter output current, are determined by using artificial neural networks. Neuro-fuzzy controlled and PI controlled inverter simulation results are given together and compared. Simulation results show that grid interactive inverter operates synchronously with the grid line; inverter output current which imported to the grid is in sinusoidal waveform and harmonic level of this current meets the international standards. Also, it is seen that inverter has faster response to the reference variations than PI controller.