Solar Radiation Prediction using Machine Learning Model

• Published in: 2023 International Conference on Sustainable Computing and Data Communication Systems (ICSCDS) pp. 52 - 57
• Main Authors: Venu, K., Indhu Prakash, K. V., Jayaram, S., Karan, N. S., Raja, M.Manicka, Renu, K.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 23.03.2023
• Subjects: Boosting; Global Horizontal Irradiance (GHI); Gradient Boosting (XG Boost); Long Short-Term Memory (LSTM); Measurement; Prediction; Predictive models; Solar Radiance Predictions; Solar radiation; Support vector machines; Training; Weather forecasting
• DOI: 10.1109/ICSCDS56580.2023.10104904

The two primary categories of methods for predicting the world's solar radiation were scientific estimates & machine learning models. The purpose of this work is to provide a summary of solar radiation prediction in this context using machine learning algorithms. It will be shown that, despite the fact that several studies describe methodology such neural network models or support vector regression, different methods (such as regression model, random forest, XGBoost, etc.) tend to be used in this prediction. Ranking the efficiency of such methods is difficult because of the diversity of the data gathering, time step, predicting range, setup, and performance metrics. The prediction inaccuracy is quite comparable overall. To improve the effectiveness of forecasts, some authors advocated the adoption of hybrid models. The primary goals of this study are to assess the level of prediction by using meteorological information and to determine accuracy following training and testing. Results from Gradient Boosting (XGBoost) were compared to those obtained from LSTM model simulations. According to the results, LSTM network surpassed XGBoost on the same dataset by a large margin with a normalized the Root Mean Square Error (RMSE) value of 0.02%. Because of dataset is a time - series data one, LSTM performs better.