Major project
Energy, Power & Utilities
Solar plant generation forecasting and fault flagging
Forecast a solar plant's output from the weather and flag inverters that under-perform.
The brief
A solar developer runs plants in India and must schedule power with the grid a day ahead. They need a generation forecast from weather data, and an alert when an inverter produces less than it should, so a technician can be sent. Good looks like: a day-ahead forecast with known error, under-performance alerts that match real faults, and a plant dashboard.
What you will build
- Join plant generation data with weather and irradiance
- Build day-ahead forecasts and compare models
- Model expected output per inverter and flag shortfalls
- Validate alerts against visible faults in the data
- Use irradiance data to compare sites
- Build a plant operations dashboard and write the report
Free, open data to build it with
- Solar Power Generation Data (two plants in India) ↗ Kaggle
- NASA POWER: solar and weather data ↗ NASA
- Global Solar Atlas ↗ World Bank / Solargis
Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.