TUTCRIS - Tampereen teknillinen yliopisto

TUTCRIS

Knowledge-based optimization of artificial neural network topology for process modeling of fused deposition modeling

Tutkimustuotosvertaisarvioitu

Yksityiskohdat

AlkuperäiskieliEnglanti
OtsikkoProceedings of the ASME 2018 International Design Engineering Technical Conferences & Computers and Information in Engineering Conference (IDETC/CIE 2018)
AlaotsikkoVolume 4: 23rd Design for Manufacturing and the Life Cycle Conference; 12th International Conference on Micro- and Nanosystems. Aug. 26-29, 2018, Quebec City, Canada.
KustantajaASME International
Sivumäärä12
Vuosikerta4
ISBN (elektroninen)978-0-7918-5179-1
DOI - pysyväislinkit
TilaJulkaistu - 2018
OKM-julkaisutyyppiA4 Artikkeli konferenssijulkaisussa
TapahtumaASME International Design Engineering Technical Conferences & Computers and Information in Engineering Conference -
Kesto: 1 tammikuuta 2000 → …

Conference

ConferenceASME International Design Engineering Technical Conferences & Computers and Information in Engineering Conference
Ajanjakso1/01/00 → …

Tiivistelmä

Additive manufacturing (AM) continues to rise in popularity due to its various advantages over traditional manufacturing processes. AM interests industry, but achieving repeatable production quality remains problematic for many AM technologies. Thus, modeling the influence of process variables on the production quality in AM can be highly beneficial in creating useful knowledge of the process and product. An approach combining dimensional analysis conceptual modeling, mutual information based analysis, experimental sampling, factors selection, and modeling based on knowledge-Based Artificial Neural Network (KB-ANN) is proposed for Fused Deposition Modeling (FDM) process. KB-ANN reduces the excessive amount of training samples required in traditional neural networks. The developed KB-ANN’s topology for FDM, integrates existing literature and expert knowledge of the process. The KB-ANN is compared to conventional ANN using prescribed performance metrics. This research presents a methodology to concurrently perform experiments, classify influential factors, limit the effect of noise in the modeled system, and model using KB-ANN. This research can contribute to the qualification efforts of AM technologies.

Julkaisufoorumi-taso

Tilastokeskuksen tieteenalat