Showcase: Wind Energy Forecasting and Action by FogHorn
Previous
Next
The solution provides the ability to generate alerts at least 90-minutes ahead of time for turbines that cannot meet 24 hours energy generation forecast and apply it to either optimize turbine settings or revise forecast.
Key technical capabilities
- 6 month+ historical and real-time data collected from SCADA controller system on each turbine, augmented with weather, atmosphere and terrain data
- Train models based on 20+ attributes to predict power generation over 15-min intervals
- Apply models at the edge to produce real-time scores on power generation. VEL expressions are used to compare performance against forecast to generate alerts
- Enable technicians to either fine tune turbine settings or revise forecast
Architectural diagram / data flow chart
Technologies Used
- SAP HANA
- Vel&CEP
- Cloud Foundry
- Docker
- Historian
- MQTT
- ML