SunGrid!

Manage and analyze solar energy potential

SunGrid uses XGBoost and practical AI analysis to provide clear, dependable solar irradiance forecasts. Built for energy management, it translates real weather data and proven machine learning into actionable operational insights.

View the project on GitHub.

☀️ Weather Inputs

Date/time features auto-calculated.

Predicted Global Horizontal Irradiance (GHI)
— W/m²
Submit inputs to get started.

🔍 Feature Importance

Features ranked by predictive power. Higher scores indicate greater influence on solar irradiance predictions.

# Feature Name Importance Score
1 Month 4684.58
2 Wind Direction 3271.83
3 Day 2841.93
4 Temperature 1651.69
5 Humidity 1588.09
6 Speed 765.86
7 Pressure 523.79

📊 Model Performance

Dataset: 32,686 meteorological observations (4 months).
Train/Test split: 26,148 / 6,538 samples.

R² Score
(Goodness of Fit)
0.93
MAE
(Error Magnitude)
33.19
RMSE
(Outlier Sensitivity)
82.99