With the LoadManager|Forecast you may forecast time series such as energy consumption, energy generation and prices for very varied timescales – from years and months to the day ahead and intraday periods. The forecasts can be carried out manually or automatically in jobs via the graphical user interface. The LoadManager|Forecast supports you in developing forecast scenarios and the creation of ensemble forecasts.
Our experts have developed models which can take into account the accompanying influencing factors as well as the measured energy time series. The LoadManager|Forecast incorportaes both our know-how and experience from many years of practical applications amongst energy suppliers and in industry. The forecast models work on the basis of:
- Comparison days
- Average value procedures
- Function approximation
- Artificial neural networks
Apart from standard models preconfigured by our experts, the LoadManager|Forecast also has models which may be freely configured by the users. Optimisation algorithms can set up the model suitable for your tasks and implement it into daily practical use.
We have developed special forecast models for the requirements of energy suppliers. Thanks to the very high forecast quality the need to use balancing energy is avoided. Due to the high performance, you can use the LoadManager|Forecast to individually forecast customers with measured power consumption or do so as a group. Standard load profiles are used to forecast the consumption of non-measured customers. All normal profile procedures are available for this – with and without temperature dependency.
Forecast objects are defined in topologies. Here you have the opportunity to create the space for the future forecast customer load curves according to their allocation for grid zone, balancing grid, distribution grid and supplier. If required you can automatically set up forecast topologies from the allocation lists which you receive within the framework of GPKE/GeLi Gas exchange processes. You thus also have a firm grip on your sales forecasts when there are customer changes.
Grid operators forecast the grid loads of partial grids right up to the whole grid using the LoadManager|Forecast. In this, they can use a top-down or bottom-up approach, or mixtures of the two. For non-load-measured customers the standard load profile library of the LoadManager|Forecast is available. It contains both analytical and synthetic approaches for the forecast.
You can automatically recreate the structure of your grid area in topologies by using customer information lists. This provides a solid data basis for forecasting the difference balancing group. Additionally, load forecasts for different grid points may be calculated for grid management.
Electricity generation from wind, solar and hydropower fluctuates in dependency on the weather. This poses large challenges for both direct sellers and grid operators. In order to meet these challenges we have developed optimised forecast models which, apart from weather forecasts, also take into account the geographical conditions and turbine specifications.
If required, we supply you with the underlying high-definition weather data (measurements and forecasts) as a service with up to four updates per day. Via interfaces this data is automatically imported into the LoadManager|EDM and is then available for EEG infeed forecasts.
You can develop and optimise the infeed forecast models yourself using the LoadManager|Forecast. Shutdowns and repowering are treated separately when training the models in order to obtain the best possible forecast quality. Within the framework of our service offer you may call on our experts for the model training if required. We can insert preconfigured models directly into your LoadManager database – either at intervals or initially.