Forecasting

What is Forecaster:

Based on the AIoT platform EnOS™, Envision Digital Renewable Power Forecast product has integrated the world leading numerical weather forecast data from ECMWF/Met Office/NOAA/DWD/CMC/IBM and 4D actual observation data from renewable site/radar/satellite to establish a refined 1kmX1km weather forecast model customized especially for renewable energy on HPC. And the Forecast product intends to provide the best renewable power forecast according to combine the best numerical weather forecast and the unique power output characteristics of each turbine and inverter through AI technology.

Why Forecaster?

1. Expertise in meteorology

Global R&D capability: Envision has established Aerodynamics and Meteorology R&D Center in Singapore, where over 40 experts specialized in meteorology and AI are working on improving the weather forecast technology and renewable power forecast capabilities. Besides in-house weather forecast model, Envision has good research relationships with the world leading weather forecast center, including ECMWF/NCAR/IBM. So, the wind and solar weather forecast accuracy can be improved by 30%.

2. Skilled at wind turbine

Based on 100GW of renewable energy asset management experience, the Forecast product analyzes the production characteristics of each type of wind turbine by using physical and advanced machine learning algorithms to forecast turbine level power output, which can decrease the deviation by 10%.

3. Experienced in PV

By combining rich industry experience, massive meteorological data and machine learning, Envision has built physical digital models of PV power station and the inverter to analyze numerical weather forecast and the energy conversion of PV. Through these models, we can identify the effects of temperature, dust, and shadow on power generation efficiency and finally reduce the deviation by 10%.

4. High stability

  • Various kinds of data transfer
  • Real-time alarm
  • Hot standby
  • Maintenance guarantee

5. Value Creation

  • Integrate global resources to provide more accurate renewable power forecast to meet Grid’s need.
  • Customized medium and long term meteorological services help the annual production plans and large component maintenance plans, etc.
  • Provide more accurate power generation forecast to give strong support for power trading market.

6. Data Accuracy

Accuracy over 88% for wind power forecast
Accuracy over 93% for PV power forecast
5% better than industry average accuracy
Reduce 50% of penalties associated with inaccurate forecasts

7. Key features

1. Regional centralized power forecast services for wind and solar, including:

  • 4 hours ahead power forecast time resolution: 15 min   time span: 4 h ahead
  • Day ahead power forecast time resolution: 15 min   time span: day ahead
  • Week ahead power forecast time resolution: 15 min   time span: week ahead
  • Month ahead power forecast time resolution: 1 day time span: month ahead
  • Year ahead power forecast time resolution: 1 month  time span: year ahead

2. Working window suggestion based on weather and power forecast

3. Forecast performance evaluation and penalty calculation