Optimal management of hybrid PV/fuel cell/battery power system: A comparison of optimal hybrid approaches

In this paper, different optimal hybrid techniques have been proposed for management of a hybrid power generation system including photovoltaic (PV),. fuel cell and battery. The main power of the hybrid system comes from the photovoltaic panels, while the fuel cell and batteries are used as back up units. In order to achieve maximum power point tracking for the photovoltaic system, both fuzzy logic controller and perturb and observation methods are examined and their performances have been investigated via simulations. Next, the performance of the hybrid system has been improved via employing a family of well-known optimization approaches for load sharing among the available resources. Imperialist Competitive Algorithm (ICA), Particle Swarm Optimization (PSO), Quantum behaved Particle Swarm Optimization (QPSO), Ant Colony Optimization (ACO), and Cuckoo Optimization Algorithm (COA) are used to manage the load sharing to achieve optimal performance while the system constraints are met. The optimal performance has been characterized via the control strategy performance measure being the ratio of the amount of hydrogen production with respect to the hydrogen consumption. In order to verify the system performance, simulation studies have been carried out using practical load demand data and real weather data (solar irradiance and air temperature). Different combination of maximum power point tracking methods with various optimization algorithms have been compared with each other. The results show that the combination of fuzzy logic controller with QPSO has the best performance among the considered combinations. In this situation, when the solar irradiation is noticeably high, the required load is supplied mainly by PV array, while the battery is charged, simultaneously. In the other times, the load is mainly fed by the battery and fuel cell while the performance constraints of battery is met and the daily performance measure is optimized. (C) 2014 Elsevier Ltd. All rights reserved.