SuryaGrid AI
Real-Data Solar Forecasting & DSM Risk Engine
Phase 1.5 · Real weather (Open-Meteo), Kaggle-trained ML forecasting, an advanced configurable DSM engine, substation data, and a fully sourced multi-agent pipeline.
Open DashboardSolar Nowcasting
pvlib physics forecasts from real GHI/DNI/DHI irradiance, cloud cover, and temperature.
DSM Penalty Engine
Deviation Settlement analysis against scheduled MW with penalty cost estimates.
Energy Balance
Production vs consumption, surplus/deficit, self-consumption and grid flow.
Settlement Engine
Reward, penalty and discount settlement between owners and consumers.
RL Optimization
Reinforcement learning tunes rates, trained on real historical irradiance.
Live Real Data
Open-Meteo irradiance, persisted in a real database, no API keys required.