CHARACTERIZATION OF PHOTOVOLTAIC YIELDS IN SOUTHERN REGION OF NIGERIA (A CASE STUDY OF DELTA STATE)

Abstract

This study examines the characterization of photovoltaic yields in Delta State, Nigeria, using Artificial Neural Networks (ANN) to analyze and predict solar energy production. This study utilized experimental data to assess the impact of environmental factors such as sunlight intensity, humidity, and temperature on solar panel efficiency. The ANN was trained with data collected over 365 days to predict photovoltaic outputs. Compared to traditional modeling techniques, the ANN model demonstrated superior accuracy and robustness, achieving a coefficient of determination (R²) of 99.9%. The study underscored the potential of ANN in minimizing the need for costly physical experiments by accurately simulating and predicting photovoltaic yields based on historical data. Key findings revealed seasonal variations in photovoltaic outputs, with lower yields associated with cooler temperatures and higher rainfall. These insights are vital for optimizing the design and deployment of solar energy systems in tropical climates, highlighting the practical applications of ANN in enhancing renewable energy solutions. This study, contributes significantly to the field by providing a reliable, cost-effective method for improving solar energy efficiency and guiding future technological and policy decisions in the energy sector.