elib
DLR-Header
DLR-Logo -> http://www.dlr.de
DLR Portal Home | Impressum | Datenschutz | Barrierefreiheit | Kontakt | English
Schriftgröße: [-] Text [+]

ADVANCED GENERATIVE NEURAL NETWORKS FOR PREDICTING COMPLEX 2D PHYSICAL FIELDS WITH MINIMAL DATA: A VQVAE-TRANSFORMER FRAMEWORK

Schmitz, Andreas und Schaffrath, Robert und Voß, Christian und Karimian, Aryan und Singh, Deeksha und Heinen, Dominik (2025) ADVANCED GENERATIVE NEURAL NETWORKS FOR PREDICTING COMPLEX 2D PHYSICAL FIELDS WITH MINIMAL DATA: A VQVAE-TRANSFORMER FRAMEWORK. In: 70th ASME Turbo Expo 2025: Turbomachinery Technical Conference and Exposition, GT 2025. ASME Turbomachinery Technical Conference and Exposition, 2025-06-16 - 2025-06-20, USA. doi: 10.1115/GT2025-151756.

Dieses Archiv kann nicht den Volltext zur Verfügung stellen.

Kurzfassung

This paper introduces a novel generative neural network system for predicting complex, structured 2D physical fields with a focus on achieving high accuracy, stability, and computational efficiency with very limited data. The proposed framework employs a Vector Quantized Variational Autoencoder (VQVAE) combined with a generative model that integrates a unique combination of transformer-based structures, inverted ResNet modules, and mobileSR architectures. This architecture does not exist in its current form and has been specifically optimized for the challenges of turbomachinery applications, where data availability is often limited to less than 100 samples. The system addresses this challenge by minimizing overfitting while maintaining robustness, demonstrating its potential for future use in industrial design and optimization processes. The capabilities of the proposed system are evaluated through two different case studies: a high-temperature heat pump radial compressor, which is the second stage in a three-stage compression system for superheated steam, and the first stage of a three-stage axial compressor used in the military sector, where the rotor geometry is varied. These case studies illustrate the model’s applicability across different technical domains, highlighting its potential to serve as a versatile tool in simulationdriven design processes. We analyze the model’s architecture and provide detailed insights into its ability to produce physically meaningful predictions, thereby demonstrating its utility in scenarios that traditionally rely on computationally intensive simulation models. While the model is capable of predicting 0D performance metrics such as efficiency—similar to traditional Kriging-based surrogate models—the authors argue that this should not be considered its primary application. The benefit of the model here will be very limited. Unlike traditional surrogate modelling, which reduces design variables to scalar values, this model directly predicts full 2D flow fields, providing significantly richer information. As a result, optimization processes should be rethought to take advantage of this additional level of detail. Rather than treating the model as a simple function approximator for efficiency or pressure loss, engineers can leverage its spatially resolved predictions to directly identify and address local flow phenomena. The fast 2D predictions allow for the identification of undesirable flow patterns that remain invisible in traditional 0D metrics

elib-URL des Eintrags:https://elib.dlr.de/216378/
Dokumentart:Konferenzbeitrag (Vortrag)
Titel:ADVANCED GENERATIVE NEURAL NETWORKS FOR PREDICTING COMPLEX 2D PHYSICAL FIELDS WITH MINIMAL DATA: A VQVAE-TRANSFORMER FRAMEWORK
Autoren:
AutorenInstitution oder E-Mail-AdresseAutoren-ORCID-iDORCID Put Code
Schmitz, AndreasAndreas.Schmitz (at) dlr.deNICHT SPEZIFIZIERTNICHT SPEZIFIZIERT
Schaffrath, RobertRobert.Schaffrath (at) dlr.dehttps://orcid.org/0000-0001-8487-8299NICHT SPEZIFIZIERT
Voß, ChristianChristian.Voss (at) dlr.dehttps://orcid.org/0009-0007-0504-495X192442273
Karimian, Aryanaryan.karimian (at) dlr.deNICHT SPEZIFIZIERTNICHT SPEZIFIZIERT
Singh, Deekshadeeksha.singh (at) dlr.deNICHT SPEZIFIZIERTNICHT SPEZIFIZIERT
Heinen, Dominikdominik.heinen (at) dlr.deNICHT SPEZIFIZIERTNICHT SPEZIFIZIERT
Datum:2025
Erschienen in:70th ASME Turbo Expo 2025: Turbomachinery Technical Conference and Exposition, GT 2025
Referierte Publikation:Ja
Open Access:Nein
Gold Open Access:Nein
In SCOPUS:Ja
In ISI Web of Science:Nein
DOI:10.1115/GT2025-151756
Status:veröffentlicht
Stichwörter:Generative neuronale Netze, VQVAE, Turbomaschinen, Strömungsfeldvorhersage, datenarme Szenarien, Surrogatmodellierung, Designoptimierung
Veranstaltungstitel:ASME Turbomachinery Technical Conference and Exposition
Veranstaltungsort:USA
Veranstaltungsart:internationale Konferenz
Veranstaltungsbeginn:16 Juni 2025
Veranstaltungsende:20 Juni 2025
HGF - Forschungsbereich:keine Zuordnung
HGF - Programm:keine Zuordnung
HGF - Programmthema:keine Zuordnung
DLR - Schwerpunkt:Digitalisierung
DLR - Forschungsgebiet:D CPE - Cyberphysisches Engineering
DLR - Teilgebiet (Projekt, Vorhaben):D - HyOpt
Standort: Köln-Porz , Zittau
Institute & Einrichtungen:Institut für Antriebstechnik > Fan- und Verdichter
Institut für CO2-arme Industrieprozesse > Hochtemperaturwärmepumpen
Hinterlegt von: Schmitz, Andreas
Hinterlegt am:22 Sep 2025 20:03
Letzte Änderung:22 Sep 2025 20:03

Nur für Mitarbeiter des Archivs: Kontrollseite des Eintrags

Blättern
Suchen
Hilfe & Kontakt
Informationen
OpenAIRE Validator logo electronic library verwendet EPrints 3.3.12
Gestaltung Webseite und Datenbank: Copyright © Deutsches Zentrum für Luft- und Raumfahrt (DLR). Alle Rechte vorbehalten.