06 header e

Gallery

New forecasting algorithm for photovoltaic generation

For the operation of autarkic energy systems, the most accurate prediction possible of the expected generation is eminent for stable and optimized operation. As part of the research carried out in IREN2, measurements were taken at various photovoltaic plants and the forecasts compared with the actual course. Different prognosis procedures were used.

In the research project, methods were developed that could greatly improve the quality of previous forecasts. In doing so, effective radiation values are calculated by mathematical methods taking into account all geometrical properties of the plant and geography of the location, which better map the course of production over the day.

The results are impressive - the forecasts have been improved by 40% and achieve an average accuracy of more than 90% - more than 95% on sunny days. Since mid-November 2017, these methods have been successfully implemented in the IREN2 system in Wildpoldsried.

consortium

AÜW AllgäuNetz Siemens RWTH ID.KOM HS Kempten