Doctoral thesis (Dissertations and theses)
Data Driven Surrogate Frameworks for Computational Mechanics: Bayesian and Geometric Deep Learning Approaches
DESHPANDE, Saurabh
2023
 

Files


Full Text
Thesis_Saurabh_Deshpande.pdf
Author postprint (56.03 MB)
Download
Annexes
defense_presentation_Saurabh.mp4
(235.8 MB)
Download
defense_slides_Saurabh.pdf
(41.07 MB)
Download
references.bib
(237.95 kB) Creative Commons License - Attribution
Download

All documents in ORBilu are protected by a user license.

Send to



Details



Keywords :
Deep learning, data driven modelling, computational mechanics
Abstract :
[en] In modern applications, high-fidelity computational models are often impractical due to their slow performance and also lack information about the certainty of their predictions. Deep learning techniques have recently emerged as a powerful tool for accelerating such predictions. However, these techniques can be inefficient when confronted with larger and more complex problems. This thesis introduces innovative deep learning surrogate frameworks that are scalable, robust, require minimum hyper-parameter tuning, are fast at the inference stage, and are accurate in forecasting non-linear deformation responses of solid objects. These surrogate frameworks are constructed using various deep learning techniques under deterministic as well as Bayesian settings. Bayesian frameworks enable us to capture uncertainties and provide a means to trust the predictions of the neural network approaches. This thesis introduces a new geometric deep learning framework, called MAgNET (Multi-channel Aggregation Network). MAgNET is designed to handle large-dimensional graph-structured data using an encoder-decoder architecture. MAgNET is built upon the novel MAg (Multichannel Aggregation) operation, which generalises the concept of multi-channel local operations found in convolutional neural networks to arbitrary non-grid inputs. The MAg layers are combined with the novel graph pooling/unpooling operations to form a powerful graph U-Net architecture capable of efficiently performing supervised learning on large-dimensional graph-structured data, like complex meshes. Additionally, the thesis demonstrates the use of state-of-the-art attention-based networks, which have revolutionized various engineering fields but have remained unexplored for their uses in the field of computational mechanics. We demonstrate the efficiency and versatility of the proposed frameworks by applying them to surrogate modeling for non-linear finite element simulations. Our suggested methods, particularly the MAgNET architecture, possess broad applicability, enabling researchers and practitioners to explore novel modeling scenarios and applications. Through the open sharing of the source codes and datasets employed, this thesis not only makes a significant contribution to the field of surrogate modeling in mechanics but also paves the way for numerous research opportunities for their utilisation in various engineering and scientific applications.
Research center :
ULHPC - University of Luxembourg: High Performance Computing
Disciplines :
Engineering, computing & technology: Multidisciplinary, general & others
Author, co-author :
DESHPANDE, Saurabh  ;  University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Engineering (DoE)
Language :
English
Title :
Data Driven Surrogate Frameworks for Computational Mechanics: Bayesian and Geometric Deep Learning Approaches
Defense date :
18 September 2023
Institution :
Unilu - University of Luxembourg [Faculty of Science, Technology and Medicine], Luxembourg
Degree :
Docteur en Sciences de l'Ingénieur (DIP_DOC_0005_B)
Jury member :
BORDAS, Stéphane ;  University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Engineering (DoE)
ZILIAN, Andreas  ;  University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Engineering (DoE)
LENGIEWICZ, Jakub ;  University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Engineering (DoE) ; PAN - Polish Academy of Sciences [PL]
Cotin, Stéphane;  INRIA > Research Director
Chatzi, Eleni;  ETH Zürich
European Projects :
H2020 - 764644 - RAINBOW - Rapid Biomechanics Simulation for Personalized Clinical Design
Name of the research project :
R-AGR-3325 - H2020-MSCA-ITN-2017-764644-RAINBOW (01/04/2018 - 31/03/2023) - BORDAS Stéphane
Funders :
European Union. Marie Skłodowska-Curie Actions [BE]
Union Européenne [BE]
Available on ORBilu :
since 07 November 2023

Statistics


Number of views
138 (42 by Unilu)
Number of downloads
135 (17 by Unilu)

Bibliography


Similar publications



Contact ORBilu