MULTI-MODAL ENERGY PREDICTION FOR A SUSTAINABLE ENERGY FUTURE

This study utilizes machine learning methods, Random Forest Regression and Linear Regression to predict wind and solar energy generation, to enable prediction using historical weather and energy data. Predicting solar power was successfully achieved with Random Forest Regression, yielding an impressive accuracy (R2 = 96.1%) thanks to its ability to seize complex and non-linear relationships, while Linear Regression delivered feasible predictions of wind power (R2 = 78.4%), attributed to its linear dependence on wind speed. These models strike a trade-off between high accuracy and computational efficiency, allowing them to be used for real-time and short-term forecasting. Such work will further integration of renewable energy, stability in the grid, and optimal use of energy, as we develop scalable and adaptable solutions to the challenges of sustainable energy systems. Advanced algorithms and seasonality analysis will improve long-term prediction accuracy in future developments.