Ultra-fast generation forecasting will improve the energy management of photovoltaic farms

Enercode has signed a contract with Tauron EkoEnergia for the project “Ultra-fast generation forecasting and control of photovoltaic panels”.

Algorithms from the field of artificial intelligence and machine learning are popular tools used in forecasting. Commercially available RES forecasts are based on meteorological data from numerical weather prediction models with a spatial grid resolution of several kilometers (Thor).

By their very nature, such forecasts are subject to error, as they cannot predict e.g. the exact location of clouds that affect generation levels. For predictions carried out for commercial balancing purposes, such an error is generally of minor importance, but it becomes significant in local balancing issues, where, for example, in order to stabilize the grid’s operating parameters, the operating points of controllable units have to be changed within a minute, so as to balance out energy shortages, e.g. from wind or solar power. In such cases, you need a much more accurate forecasts, which are able to use data such as local meteorological stations, or images of the sky. Enercode’s task is to provide the most accurate energy forecast using the assimilation of local measurement data.

The contract signed on 14.06.2021 provides that the cooperation will last 12 months and will be carried out in collaboration with the Interdisciplinary Department of Energy Analysis, NCBJ.