Improvement of the energy production of a photovoltaic-wind hybrid system using NF-PSO MPPT

Abstract

This manuscript gives a contribution to the optimization of a hybrid Photovoltaic-Wind Turbine system with a storage system. In order to capture the maximum power that can be produced by each source, while maintaining the rotor speed of the wind turbine at its maximum values according to wind variations, the Neuro-Fuzzy-Particle Swarm Optimization (NF-PSO) controller is proposed. The Neuro-Fuzzy method is used here because it allows an automatic generation of fuzzy rules, and the Particle Swarm Optimization to find an optimal gain allowing to readjust the dynamics of the fuzzy rules by reducing the power losses (oscillations). For the proper functioning of such a system, we have developed a fuzzy supervisor in order to have an optimal control of the system according to the variations of the requested load and the produced power by considering the storage system and the load shedding. The simulation results of the system confirmed the better performance of this method in terms of speed with a response time of 0.2s on the wind side and 0.025s on the side photovoltaic, of efficiency with 99.87% on the photovoltaic side and 99.6% on the wind side, and above all in term of oscillation reduction with practically a negligible oscillation rate compared to the NF and the Cuckoo algorithm.

Keywords

• neuro-fuzzy
• particle swarm optimization
• cuckoo
• supervisor
• oscillations