Prices, forecasts and optimization

Overview · Dashboard · Privacy

GridFusion combines current measurements with electricity prices, weather and PV forecasts. Its local Energy Engine remains independent of individual data providers.

Electricity prices

The selected provider and bidding zone determine the available wholesale intervals. Your tariff adds taxes, levies, network fees and contractual mark-ups. GridFusion supports variable interval lengths and negative prices.

If valid prices are unavailable, optimization falls back safely to self-consumption. A provider failure never blocks local metering or device operation.

Weather and solar forecast

An installation may contain several physical PV arrays with different orientation, tilt and capacity. Provider data is validated, cached and reduced to a generic PV forecast. Stale or implausible data is marked and is not treated as a current measurement.

Learning period

At first, the system does not yet know the installation’s individual production and consumption patterns. The dashboard and portal therefore indicate a learning period. A longer, high-quality history can improve forecasts for the individual device and site.

Server-side AI

Control 7 performs no AI training or inference on the device. With explicit permission, it sends a bounded set of normalized energy data. The central service calculates forecasts or recommendations and returns normalized results only.

AI cannot activate automatic hardware control, alter a safety limit or bypass the local write permission. Without valid local measurements, no new control action is executed.