WIND AND SOLAR ENERGY PREDICTION using ML/DL models

Table of Contents

• Solar energy is one of several sustainable sources that is becoming increasingly important in the energy sector due to its potential to cut carbon emissions and counteract growing electricity prices. The main issue with solar energy is that it cannot be used due to the constantly shifting and unpredictable weather, cloud cover, climate, and seasons. As a result, solar energy generation varies. Therefore, resource planners and businesses are looking for models that take these uncertainties into account for the daily design and management of solar energy production, which could enable them to meet consumer demand and supply regardless of weather conditions.
• Weather complexity makes accurate synthesis of wind output difficult, and commercial confidentiality means that historical data is frequently limited. We present and validate a model for simulating the hourly power output of wind farms located anywhere in the world.

Predicts real-time solar and wind energy. Accuracy, mean square error, root mean square error, mean absolute error are calculated for each algorithm and different hyperparameters and the best one is chosen.

Laptops, Desktops and Tablet PCs

• Frontend: HTML, CSS, Bootstrap
• Backend: JavaScript, Python
• Prediction models: ML/DL models
• ML Models: Lasso Regression, Ridge Regression, Decision Tree, SVM, Random Forest
• DL Model: LSTM

Data was collected from https://open-power-system-data.org/ which is a free open source platform with data on power systems for 37 European countries. But we chose to focus on a specific country, Germany, due to having the highest proportion of renewable energy than any other country(about 46 percent of its energy come from solar, wind, biomass) and hence it is a good indicator of where the rest of the world is headed.