Artificial Neural Network Based Solar Prediction in Multi-Area Networks

Abstract

A clean, readily accessible and renewable energy source is solar energy. The use of photovoltaic panels is a popular trend to harness solar energy for the production of electrical power. Variable generated electricity is a result of solar energy's intermittent nature. An energy management system with a solar forecast module is presented in this study. The main purpose of this study is to apply ANN for verifying and predicting solar energy in India. The output is controlled by an ANN controller. To meet the demands put forward by the Energy Management System (EMS), artificial neural networks (ANNs), a subset of statistical approaches, were selected as the prediction method. Sun rising and sun set times are included in the total of 10 input factors taken into account for solar prediction data. For the EMS to operate properly, there are needs that include accurate forecasting in the short-term prediction horizon. The Mean Square Error (MSE) parameter is used to measure performance, and all findings are implemented in the Python language. This study shows the low RMSE value of that particular area and ANN is an efficient method for this study.