Employing a Gene Expression-based Technique to Improve the Accuracy of Estimating the Total Generated Power by Neighboring Photovoltaic Systems

The penetration of photovoltaic systems (PVs) is increasing and they are considered attractive options for electricity generation in distribution networks. This study focuses on estimating the total power generated by a group of neighboring PVs, spread in a locality while using a single pyranometer for measuring the solar irradiance. A new model has been proposed which employs the Gene Expression Programming (GEP) technique for developing a correlation between the distribution of the PVs and the irradiance measured by the pyranometer in estimating the total power generated. The proposed technique considers the geographic variability reduction and employs a Wavelet Transform technique for the calculations. The effective performance of the proposed model is validated using the real data collected by the Solar Project at the University of Queensland, Brisbane, Australia. The studies reveal that the proposed technique yields more accurate results against the other existing approaches.