Modelado predictivo del potencial de generación de energías renovables en función de variables geográficas y climáticas en una región específica.

Abstract

Abstract: This thesis develops a predictive model of renewable energy generation potential, based into geographical and climatic variables in Ecuador. The main objective was to develop a model capable of estimating energy generation potential based on characteristics such as altitude, solar radiation, and temperature in a specific region of Ecuador. Open data sources obtained from Meteostat and NASA Power were used, formatted as GeoJSON, and processed using supervised learning techniques in Python and the Random Forest Regressor algorithm. The results show the solar, wind, and hydroelectric potential of each province in the country based on temporal climatic parameters such as the month of the year and the most relevant variables of the study, such as average temperature, wind speed, precipitation, and relative humidity. The model was evaluated using metrics such as Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and the coefficient of determination R². Finally, an API was generated using FastAPI and a web application using Streamlit, provided by the Scikit-learn library, with versioning available on GitHub.