Improved calculation of PV power generation in Sweden
High latitude and low electricity price in Sweden make accurate prediction of photovoltaic (PV) system energy yield fundamental.
Project manager at MDU
More precise estimates would give smaller uncertainties in economy, resulting in a more resource efficient PV development. This would benefit different investors like municipalities, building companies, housing associations and private persons.
To have access to accurate data for validation of PV software, a PV system network is created at three SMHI weather stations with complete irradiance measurements. To cope with the lack of complete data at other stations, this project aims to utilize cutting-edge machine learning techniques to generate missing data fundamental for the uses of several software tools. The best PV simulation tool will be recommended. Future foreseen solar radiation for Sweden will be performed to quantitatively analyze the effects of likely climate changes. A map of Swedish PV energy yield is produced, based on STRÅNG irradiance data.