High Efficient and Low Cost MPPT Technique of Photovoltaic system based on ANNs

In this paper, a new maximum power point tracking (MPPT) method based on two artificial neural networks (ANNs) is used to extract the maximum power of photovoltaic (PV) array. The first ANN is used to estimate the solar irradiation (G) and photovoltaic temperature (T) by using current and voltage sensors instead of using Pyranometer and temperature sensor. This estimation of G and T provides both high accuracy and low cost. The second ANN is used to estimate the desired duty cycle to harvest the maximum power for different G and T for certain loading conditions. The variations of loading are also considered in this study. Real data from a project in Hurghada city (in Egypt) is used to verify the proposed MPPT method. Comparisons with some conventional MPPT methods, e. g., perturb & observe and incremental conductance methods, are also presented.