![]() | Only 14 pages are availabe for public view |
Abstract The PV system characteristic is non-linear whose its output power varies as function of the irradiance and temperature. There are two ways to increase the output power from the photovoltaic systems, either by directing the panel to be perpendicular to the solar radiation most of the time or by extracting the optimal output power by using a maximum power point tracking control. Hence, it is essential to operate the PV system at its maximum power point. The electric energy that is produced from Photovoltaic (PV) systems depends on weather conditions, it is important to obtain the maximum power point of the array appropriately, which can be done by designing an efficient control scheme for a DC/DC converter. The controller must force the system to operate at that point, where the PV module can produce its maximum output voltage and current.In this thesis two MPPT algorithms will be investigated: first developing the Perturb & observe MPPT technique by modifying the fixed step size perturb voltage to be variable and depend on the rate of power change. Second a multilayer feed forward neural network (MFFNN) based MPP tracker for PV systems is also proposed. The proposed MFFNN is trained using back propagation algorithm and suitable training data. The MFFNN uses the irradiance and temperature as inputs. A program was designed and developed by the MATLAB /SIMULINK to generate simulation data for PV at various conditions to train and test the proposed neural network. Finally a comparison between the MFFNN based MPP tracker and the developed P&O algorithm with the convention and P&O algorithm were done. The results show that the proposed MFFNN has high accuracy, speed and didn’t affected by the sudden change of weather conditions. And lead to more stability of output around the MPP. |