Solar energy: Improved data mining tool for hourly solar radiation

“The harnessing and use of solar energy will continue to gain relevance, particularly when viewed against the background of the elevated cost of fossil fuels and their negative impact upon the environment”, Tamer Khatib (Institute of Networked and Embedded Systems) explains. Together with his colleague Wilfried Elmenreich he has developed a new approach for improved data mining for hourly solar radiation.

Elmenreich goes on to say:“Solar radiation data provide information on how much of the sun’s energy strikes the Earth’s surface at a specific location during a defined time period”. These data are needed for effective research into solar energy utilization. Due to the cost and difficulty involved in obtaining solar energy measurements, these data are not readily available; therefore, researchers have explored alternative ways of generating these data.

Khatib elucidates further: “On the one hand, there are regions in the world, for which there are no solar radiation measurement data. Therefore, a tool is required, in order to assess the potential for solar energy. On the other hand, there are regions with daily averages of solar radiation data. These are less suitable for the evaluation of solar energy systems than solar radiation data that is generated on an hourly basis.” Researchers have been working on the development of smart prediction techniques, which can extrapolate an hourly average value from the daily data.

The “Smart Grid Lab” in Klagenfurt has now successfully developed such a model. Supplied with a total of six different inputs − mean daily solar radiation, hour angle, sunset hour angle, date, latitude and longitude− the model calculates the mean hourly solar radiation. Wilfried Elmenreich is pleased with the results: “The results prove that the model can predict the hourly solar radiation very well, and with an accuracy of prediction exceeding that of the empirical and statistic models used so far.”