Abstract
This paper presents an enhanced maximum power point tracking approach to extract power from photovoltaic panels. The proposed method uses an artificial neural network technique to improve the fractional open-circuit voltage method by learning the correlation between the open-circuit voltage, temperature, and irradiance. The proposed method considers temperature variation and can eliminate the steady-state oscillation that comes with conventional algorithms, which improves the overall efficiency of the photovoltaic system. A comparison with the traditional and most widely used algorithms is discussed and shows the difference in performance. The presented algorithm is implemented with a Cuk converter and tested under various weather and irradiance conditions. The results validate the competitiveness of the algorithm against other algorithms.