A Neural Network-based System Identification Model to Predict Output Current and Voltage of Solar Photovoltaic Panels

Solar irradiance is the energy per unit area received by the Sun as electromagnetic radiation. It is one of the most important renewable energy sources. Photovoltaic or other solar technologies are used to generate power more accurately than direct sun irradiation. Solar irradiance research and measurement have a variety of critical applications, including forecasting power generation from solar power plants, climate modeling, and weather forecasting. This paper presents a neural network-based system identification model developed using measured parameters from solar panels with various wattage specifications, namely, 10W, 20W, 40W, and 100W. The parameters that were measured to train the ANN model for the prediction of the output current and voltage include the angle of panel orientation, panel temperature, ambient temperature, irradiance, and wattage. Several training experiments were conducted and the best ANN model produced at 500 epochs gave an accuracy of 99.81% and a loss of 0.1940. The model was deployed into an intelligent Web App that was also developed in this study. This app could be a potential tool for renewable energy engineers and researchers