Software and services
Raccourcis catalogue [EN]
Meteodyn Forecast
Wind power generation forecasting service, accurate and reliable
Forecasting wind power generation: a necessity
Wind is a renewable resource that is particularly difficult to predict. This complicates the work of grid operators, energy traders and wind farm owners or operators. Indeed, an incorrect electricity production forecast can have economic repercussions for all stakeholders.
Meteodyn has more than 15 years of experience and research in the wind engineering and meteorology sectors. Strengthened by that, Meteodyn's researchers have managed to combine weather forecasting and statistical machine learning to provide the most accurate production forecasting service possible.
Good wind generation forecasting is essential to maintain the balance between electricity consumption and production, reducing penalties and operating costs. This allows a finer forecast of the energy price and facilitates negotiation, with a more affordable cost for the final consumer.
Our energy production forecasting service automatically sends you estimations for your entire wind farm.
Forecasts
Forecast data provided by Meteodyn Forecast
- Wind energy production forecast for the entire farm :
- intraday: from 10 min to 4 hours,
- day ahead: 0 to 5 days,
- week ahead: up to 14 days.
- Wind speed and direction
- Probability distribution of production (P10, P25, etc.)
Accurate wind generation forecasts for operating wind farms
The implementation of this service ideally requires at least one year of wind farm production data as well as weather forecasting history to perform learning with our artificial neural network. Then, the system delivers the future production using the meteorological forecasts.
Wind production forecast for start-up wind farms
This service is also available for start-up wind farms having few or no production history.
For these wind farms, we exploit our expertise, by modeling your wind farm and its topography in our Meteodyn WT software. We implement the wind data from your measuring mast to generate the production histories needed for learning.
Then, the artificial neural network delivers the upcoming production of your farm by using the weather forecast.
These results give you more visibility in the start-up phases. In order to increase accuracy, we periodically perform re-learning sessions with your wind farm's actual production data.
Characteristics of the service
1
The indispensable weather forecasts
Our forecasting of weather variables is performed using the major numerical weather prediction models.
These weather forecasts are then transposed to the desired location, at turbine hub height.
For greater accuracy, we can perform a micro scale transposition with our Meteodyn WT software.
2
Artificial intelligence and machine learning
Thanks to our self-learning artificial neural network, the software autonomously modifies its way of considering input data. This way, the more data it analyzes, the more accurate it becomes: indispensable strategy for an accurate power production forecasting.
3
High availability of deliveries
The deliveries frequency of the energy production forecast is from once a day to 4 times an hour. These forecasts have a time step from a few minutes to several days, allowing you to interact with your customers and partners in real time.
Prerequisites & options
Required information
- The necessary horizon:
- intraday: from 10 minutes to 4 hours,
- day ahead: 0 to 5 days,
- week ahead: up to 14 days
- The expected delivery frequency
- The desired forecast time step
- The characteristics of the wind farm
- Historical production data of the park
- The delivery formats and support
Performance report
In addition to this wind production forecasting service, we can provide you with reports on our forecasts performances, using our Meteodyn WPA software.