Forecaster Renewables


Forecaster Renewables is the application that implements forecasting models for the generation of power from renewable sources. The application is based on time series models, regression models, neural networks and hybrid models that manage the necessary weather drivers.

Business Context

The need for accurate planning of energy generation is increasing. The context is complicated by the types of generation involved, i.e. photovoltaic, wind and pump storage hydroelectricity; these have complex behaviors which react differently according to different weather conditions. This problem impacts the energy generation area of the company, but the forecasting problem has many points in common with the forecasting activities already managed by sales portfolio forecasting.

ì4C Application

This application meets the need to provide increasingly accurate forecasts of energy from renewable sources, minimising the imbalance. This need is satisfied by the best statistical forecasting methods (time series models, linear regression models, neural networks and hybrid “Zhang” models) adapted to the specific and complex requirements of photovoltaic systems, wind farms or pump storage hydroelectricity plants. In particular, this solution focuses on the precision and accuracy of the data related to significant drivers for forecasting purposes and therefore of the data of advanced weather models, wind speed and direction at different heights and gusts of wind, temperatures, rain and climate stations connected to the plants. The combination of sophisticated models with updated information and drivers provide very accurate results, thus overcoming the forecasting problem for such complex systems. All operations can be scheduled, saving a great deal of time that can be dedicated to the analysis and study of the drivers and their trends, which can significantly modify a forecast and its accuracy.

Business Pain – Key Features

  • Plant mapping
  • Technical features mapping
  • Characteristic curve
  • Map weather stations on plant coordinates
  • Get weather forecasts
  • Specific meteo drivers for renewable technologies
  • Get plant data and measures
  • Grid operator data of relevant and non-relevant points
  • Grid operator formats for point measures
  • i4C standard for power generators
  • Get forecast data
  • Rectangular output format
  • Minimize unbalance from renewable sources
  • Specific models for photovoltaic, wind and flowing sources
  • Forecast at different aggregation levels
  • Forecast time horizon +33 +57 as defined by regulators
  • Advanced tools for experienced analysts
  • Minimize the required statistical skills
  • Automatic tools for less skilled users
  • Application usage on premise or SaaS
  • Reduce management effort
  • Complete operation scheduling
  • KPI and forecast management at any level