A Parsimonious Separated Representation Empowering PINN–PGD-Based Solutions for Parametrized Partial Differential Equations
Article dans une revue avec comité de lecture
Abstract
The efficient solution (fast and accurate) of parametric partial differential equations (pPDE) is of major interest in many domains of science and engineering, enabling evaluations of the quantities of interest, optimization, control, and uncertainty propagation—all them under stringent real-time constraints. Different methodologies have been proposed in the past within the model order reduction (MOR) community, based on the use of reduced bases (RB) or the separated representation at the heart of the so-called proper generalized decompositions (PGD). In PGD, an alternate-direction strategy is employed to circumvent the integration issues of operating in multi-dimensional domains. Recently, physics informed neural networks (PINNs), a particular collocation schema where the unknown field is approximated by a neural network (NN), have emerged in the domain of scientific machine learning. PNNs combine the versatility of NN-based approximation with the ease of collocating pPDE. The present paper proposes a combination of both procedures to find an efficient solution for pPDE, that can either be viewed as an efficient collocation procedure for PINN, or as a monolithic PGD that bypasses the use of the fixed-point alternated directions.
Files in this item
- Name:
- PIMM_JM_2024_GHNATIOS.pdf
- Size:
- 2.989Mb
- Format:
- Description:
- A Parsimonious Separated Repre ...
Related items
Showing items related by title, author, creator and subject.
-
Article dans une revue avec comité de lectureElectromagnetic field propagation inside composite materials represents a challenge where fiber-scale simulation remains intractable using classical simulation methods. The present work proposes an original 3D simulation ...
-
Article dans une revue avec comité de lecture
CUETO, Elias; FALCO, Antonio; DUVAL, Jean-Louis;
GHNATIOS, Chady;
CHINESTA SORIA, Francisco (Springer Science and Business Media LLC, 2020)
Non-intrusive approaches for the construction of computational vademecums face different challenges, especially when a parameter variation affects the physics of the problem considerably. In these situations, classical ... -
Article dans une revue avec comité de lectureREILLE, Agathe; HASCOET, Nicolas;
CUETO, Elias; DUVAL, Jean-Louis; KEUNINGS, Roland;
GHNATIOS, Chady;
AMMAR, Amine;
CHINESTA SORIA, Francisco (Elsevier Masson, 2019)
The present work aims at proposing a new methodology for learning reduced models from a small amount of data. It is based on the fact that discrete models, or their transfer function counterparts, have a low rank and then ... -
Article dans une revue avec comité de lecture
GHNATIOS, Chady; DELPLACE, Frank; BARASINSKI, Anais; DUVAL, Jean-Louis;
CUETO, Elias;
AMMAR, Amine;
CHINESTA SORIA, Francisco (Wiley, 2020)
Composite manufacturing processes usually proceed from preimpregnated preforms that are consolidated by simultaneously applying heat and pressure, so as to ensure a perfect contact compulsory for making molecular diffusion ... -
Article dans une revue avec comité de lectureIBAÑEZ, Ruben; CASTERAN, Fanny; ARGERICH, Clara; HASCOET, Nicolas; CASSAGNAU, Philippe;
GHNATIOS, Chady;
AMMAR, Amine;
CHINESTA SORIA, Francisco (MDPI, 2020)
This paper analyzes the ability of different machine learning techniques, able to operate in the low-data limit, for constructing the model linking material and process parameters with the properties and performances of ...