Abstract
SOLAR RADIATION ANALYSIS AND PREDICTION USING MACHINE LEARNING ALGORITHMS
Mohammad Hasan*, Saleh Muhammad Maruf, Abu Hena Md. Mustafa Kamal and Sabrina Sharmin
ABSTRACT
Accurate solar radiation measurement and prediction is crucial for agriculture, weather forecasting, health awareness, effective energy management, grid stability and optimizing the utilization of solar energy resources. Global temperature is significantly influenced by solar radiation since even little variations in the sun's energy output can have a big impact on the planet's climate. The existing solar radiation models measure solar radiation using physical equipments. These measurements are quite difficult, complex and costly. So, we implemented a model using machine learning algorithms for analyzing and predicting solar radiation, with the goal of enhancing the efficiency and integration of solar power systems. This project beginswith an extensive collection of solar radiation data from Kaggle.com, including weather stations, satellite imagery, and ground-based sensors. The dataset is preprocessed and we applied feature engineering techniques to extract relevant meteorological and environmental parameters. Different machine learning techniques including Linear Regression, Decision Tree Regressor, Support Vector Machines, KNN, XGBoost Regressor, Random Forests, and Gradient Boosted Algorithm, are used to train and evaluate the prepared dataset. Then, the performance of these algorithms is compared. This project will showcase the effectiveness of machine learning algorithms by predicting accurate solar radiation. The model will have the potential to significantly impact the renewable energy sector, supporting the transition to a clean and sustainable energy future.
[Full Text Article] [Download Certificate]