ForschungPublikationen
Machine learning approach for optimization of Automated Fiber Placement processes

Machine learning approach for optimization of Automated Fiber Placement processes

Kategorien Vortrag
Jahr 2017
Autoren Hocke, T., Brüning, J., Denkena, B., Dittrich, M.-A.:
Veröffentlicht in 1st CIRP Conference on Composite Materials Parts Manufacturing, 09.06.2017, Karlsruhe, 13 Seiten.
Beschreibung

Automated Fiber Placement (AFP) processes are commonly deployed in manufacturing of lightweight structures made of carbon fibre reinforced polymer. In general, AFP is connected to individual manufacturing knowledge during process planning and time consuming manual quality inspections. In both cases, automatic solutions provide a high economic potential. Therefore, a machine learning approach for planning, optimizing and inspection of AFP processes is presented. Process data from planning, CNC and online process monitoring is aggregated for the documentation of the part specific manufacturing history and the automated generation of manufacturing knowledge. Within this approach a complete automation of data capturing, data storing, modeling and optimizing is achieved.