Fraunhofer Institute for Microstructure of Materials and Systems IMWS
Fraunhofer Institute for Microstructure of Materials and Systems IMWS

AI-supported analysis of inverter efficiency: Fraunhofer IMWS receives Best Poster Award at the PV Symposium

David Daßler (vorne Mitte) mit dem Team des Fraunhofer CSP beim PV Symposium in Bad Staffelstein.
© Fraunhofer IMWS
David Daßler (front row, centre) and the Fraunhofer CSP team at the PV Symposium in Bad Staffelstein.

A team from the Fraunhofer Institute for Microstructure of Materials and Systems IMWS and partners have used data-based analyses to investigate the efficiency of inverters in photovoltaic (PV) systems. Using artificial neural networks, they were able to identify the causes of long-term yield losses. David Daßler and his team have now been awarded the Best Poster Award at the PV Symposium in Bad Staffelstein for their results.

The findings presented on the poster are based on the "robStROM" project, which was funded by the German Federal Ministry for Economic Affairs and Climate Protection. In this project, the research team evaluated the operating data of a PV system with nine central inverters for the period 2012 to 2023. Relevant values such as solar radiation, position of the sun and ambient temperature were also taken into account.

Based on the data (DC current, DC voltage, DC power, AC power and inverter efficiency), the team calculated the degradation of the efficiency curves over twelve years. The results showed long-term efficiency losses of around 2 percentage points (4 per cent relative) within the partial load range after ten years. In addition, correlations with specific voltage levels as well as the differentiation of additional degradation processes in the upstream PV generator were also demonstrated.

"I am very happy to receive the Best Poster Award. On the one hand, this recognition shows how important the role of inverters is for the performance of PV systems. On the other hand, our work clearly demonstrates the new insights that can be gained from innovative approaches to analyzing field data. I would like to thank the jury and of course the entire team," says David Daßler, scientist in the "PV Systems and Integration" group and first author of the poster. He was joined by Stephanie Malik, Dharm Patel, Robert Klengel, Carola Klute and Matthias Ebert from the Fraunhofer IMWS as well as Andreas Dietrich and Jan Spihola from DiSUN Deutsche Solarservice GmbH, Kai Kaufmann (DENKweit GmbH) and Carsten Hennig (saferay holding GmbH).

"Detailed, data-supported evaluations can make important contributions to the optimization and failure detection in PV systems. Our expertise therefore also strengthens the reliability and efficiency of solar power solutions," says Daßler.

A total of 54 poster contributions were presented at the 40th edition of the PV Symposium in Bad Staffelstein. The jury assessed these based on the degree of innovation, topicality and relevance of the content presented as well as an understandable and attractive presentation.

(March 14, 2025)