XSG-based control scheme for a grid-connected hybrid generation system

This paper presents a control scheme for the optimization of the efficiency of a grid-connected hybrid generation system consisting of a photovoltaic generator and a wind turbine. The design of the control system is made using a Xilinx System Generator tool that allows the future implementation of the code in a Field-Programmable Gate Array board. An online-trained Artificial Neural Network-based control scheme has been used in order to improve the performance of the classical control algorithms. A recurrent Elman Neural Network and a Feed Forward Neural Network have been chosen in order to maximize the power produced by the two renewable energy-based sources. Furthermore, the supervision of the grid-connected inverter is ensured by means of a traditional Voltage Oriented Control scheme. The simulation results, that have been obtained in a Matlab/Simulink environment, prove the effectiveness and the accuracy of the developed control system.