PVSwarm
RUNNING TIME:
01/2024
—
12/2026
Total project duration:
3 Years
KI generiertes Bild einer Photovoltaikanlage Bild: IntelligentVisualDesign auf Pixabay
The project
PVSwarm aims to combine swarm intelligence and machine learning to improve the condition assessment of individual photovoltaic strings. This should make it possible to distinguish between global influencing factors (e.g. weather, season, sun position) and fault sources that occur on individual strings (e.g. short circuit, ageing). This will be validated in a proof of concept. This is being done in collaboration with our partners Lakeside Labs, Campus 02, Novunex and W.I.R. Sonnen Contracting.
Our activities in the project
JOANNEUM RESEARCH's tasks in the project include, in particular, the development of machine learning methods for PV monitoring and their visual representation in a proof of concept. This also requires the coupling of swarm intelligence and machine learning methods from (partial) results. The focus in the area of machine learning is on anomaly detection methods and semi-supervised learning.
Research groups
Downloads
Customers
Project partners
Lakeside Labs GmbH
Campus 02 Fachhochschule der Wirtschaft GmbH
Novunex GmbH
W.I.R. Sonnen Contracting GmbH