Skip to main content
Erschienen in:

20.04.2024 | Original Article

A novel normalized reduced-order physics-informed neural network for solving inverse problems

verfasst von: Khang A. Luong, Thang Le-Duc, Seunghye Lee, Jaehong Lee

Erschienen in: Engineering with Computers | Ausgabe 5/2024

Einloggen

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config
loading …

Abstract

The utilization of Physics-informed Neural Networks (PINNs) in deciphering inverse problems has gained significant attention in recent years. However, the PINN training process for inverse problems is notably restricted due to gradient failures provoked by magnitudes of partial differential equations (PDEs) parameters or source functions. To address these matters, normalized reduced-order physics-informed neural network (nr-PINN) is developed in this study. The goal of the nr-PINN is to reconfigure the original PDE into a system of normalized lower-order PDEs through two sequential steps. To start with, self-homeomorphisms of the PDEs are implemented via scaling factors determined based on measured data. Afterward, each normalized PDE is transformed into a system of lower-order PDEs by primary and secondary variables. Besides, a technique to exactly impose many types of boundary conditions (BCs) by redefining NNs outputs is developed in the context of reduced-order method. The advantages of the nr-PINN model over the original one regarding solution accuracy and training cost are demonstrated through several inverse problems in solid mechanics with different types of PDEs and BCs.

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Technik"

Online-Abonnement

Mit Springer Professional "Technik" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 390 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Maschinenbau + Werkstoffe




 

Jetzt Wissensvorsprung sichern!

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

  • über 102.000 Bücher
  • über 537 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Maschinenbau + Werkstoffe
  • Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 340 Zeitschriften

aus folgenden Fachgebieten:

  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Versicherung + Risiko




Jetzt Wissensvorsprung sichern!

Literatur
1.
Zurück zum Zitat LeVeque RJ (2007) Finite difference methods for ordinary and partial differential equations: steady-state and time-dependent problems. SIAM, PhiladelphiaCrossRef LeVeque RJ (2007) Finite difference methods for ordinary and partial differential equations: steady-state and time-dependent problems. SIAM, PhiladelphiaCrossRef
2.
Zurück zum Zitat Reddy JN (2019) Introduction to the finite element method. McGraw-Hill Education, New York Reddy JN (2019) Introduction to the finite element method. McGraw-Hill Education, New York
3.
Zurück zum Zitat Logan DL (2022) First course in the finite element method, Enhanced. Cengage Learning, SI Version, Chennai Logan DL (2022) First course in the finite element method, Enhanced. Cengage Learning, SI Version, Chennai
4.
Zurück zum Zitat Raissi M, Perdikaris P, Karniadakis GE (2019) Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J Comput Phys 378:686–707MathSciNetCrossRef Raissi M, Perdikaris P, Karniadakis GE (2019) Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J Comput Phys 378:686–707MathSciNetCrossRef
5.
Zurück zum Zitat Zhuang X, Guo H, Alajlan N, Zhu H, Rabczuk T (2021) Deep autoencoder based energy method for the bending, vibration, and buckling analysis of Kirchhoff plates with transfer learning. Eur J Mech A/Solids 87:104225MathSciNetCrossRef Zhuang X, Guo H, Alajlan N, Zhu H, Rabczuk T (2021) Deep autoencoder based energy method for the bending, vibration, and buckling analysis of Kirchhoff plates with transfer learning. Eur J Mech A/Solids 87:104225MathSciNetCrossRef
6.
Zurück zum Zitat Samaniego E, Anitescu C, Goswami S, Nguyen-Thanh VM, Guo H, Hamdia K, Zhuang X, Rabczuk T (2020) An energy approach to the solution of partial differential equations in computational mechanics via machine learning: Concepts, implementation and applications. Comput Methods Appl Mech Eng 362:112790MathSciNetCrossRef Samaniego E, Anitescu C, Goswami S, Nguyen-Thanh VM, Guo H, Hamdia K, Zhuang X, Rabczuk T (2020) An energy approach to the solution of partial differential equations in computational mechanics via machine learning: Concepts, implementation and applications. Comput Methods Appl Mech Eng 362:112790MathSciNetCrossRef
7.
Zurück zum Zitat Zhu Q, Zhao Z, Yan J (2023) Physics-informed machine learning for surrogate modeling of wind pressure and optimization of pressure sensor placement. Comput Mech 71(3):481–491MathSciNetCrossRef Zhu Q, Zhao Z, Yan J (2023) Physics-informed machine learning for surrogate modeling of wind pressure and optimization of pressure sensor placement. Comput Mech 71(3):481–491MathSciNetCrossRef
8.
Zurück zum Zitat Cai S, Mao Z, Wang Z, Yin M, Karniadakis GE (2021) Physics-informed neural networks (PINNS) for fluid mechanics: A review. Acta Mech Sin 37(12):1727–1738MathSciNetCrossRef Cai S, Mao Z, Wang Z, Yin M, Karniadakis GE (2021) Physics-informed neural networks (PINNS) for fluid mechanics: A review. Acta Mech Sin 37(12):1727–1738MathSciNetCrossRef
9.
Zurück zum Zitat Li A, Zhang YJ (2023) Isogeometric analysis-based physics-informed graph neural network for studying traffic jam in neurons. Comput Methods Appl Mech Eng 403:115757MathSciNetCrossRef Li A, Zhang YJ (2023) Isogeometric analysis-based physics-informed graph neural network for studying traffic jam in neurons. Comput Methods Appl Mech Eng 403:115757MathSciNetCrossRef
10.
Zurück zum Zitat Jeong H, Bai J, Batuwatta-Gamage CP, Rathnayaka C, Zhou Y, Gu Y (2023) A physics-informed neural network-based topology optimization (PINNTO) framework for structural optimization. Eng Struct 278:115484CrossRef Jeong H, Bai J, Batuwatta-Gamage CP, Rathnayaka C, Zhou Y, Gu Y (2023) A physics-informed neural network-based topology optimization (PINNTO) framework for structural optimization. Eng Struct 278:115484CrossRef
11.
Zurück zum Zitat Wilt JK, Yang C, Gu GX (2020) Accelerating auxetic metamaterial design with deep learning. Adv Eng Mater 22(5):1901266CrossRef Wilt JK, Yang C, Gu GX (2020) Accelerating auxetic metamaterial design with deep learning. Adv Eng Mater 22(5):1901266CrossRef
12.
Zurück zum Zitat Chen Y, Lu L, Karniadakis GE, Dal Negro L (2020) Physics-informed neural networks for inverse problems in nano-optics and metamaterials. Opt Exp 28(8):11618–11633CrossRef Chen Y, Lu L, Karniadakis GE, Dal Negro L (2020) Physics-informed neural networks for inverse problems in nano-optics and metamaterials. Opt Exp 28(8):11618–11633CrossRef
13.
Zurück zum Zitat Xu C, Cao BT, Yuan Y, Meschke G (2023) Transfer learning based physics-informed neural networks for solving inverse problems in engineering structures under different loading scenarios. Comput Methods Appl Mech Eng 405:115852MathSciNetCrossRef Xu C, Cao BT, Yuan Y, Meschke G (2023) Transfer learning based physics-informed neural networks for solving inverse problems in engineering structures under different loading scenarios. Comput Methods Appl Mech Eng 405:115852MathSciNetCrossRef
14.
Zurück zum Zitat Yuan L, Ni Y-Q, Deng X-Y, Hao S (2022) A-PINN: Auxiliary physics informed neural networks for forward and inverse problems of nonlinear integro-differential equations. J Comput Phys 462:111260MathSciNetCrossRef Yuan L, Ni Y-Q, Deng X-Y, Hao S (2022) A-PINN: Auxiliary physics informed neural networks for forward and inverse problems of nonlinear integro-differential equations. J Comput Phys 462:111260MathSciNetCrossRef
15.
Zurück zum Zitat Yu J, Lu L, Meng X, Karniadakis GE (2022) Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems. Comput Methods Appl Mech Eng 393:114823MathSciNetCrossRef Yu J, Lu L, Meng X, Karniadakis GE (2022) Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems. Comput Methods Appl Mech Eng 393:114823MathSciNetCrossRef
16.
Zurück zum Zitat Gao H, Zahr MJ, Wang J-X (2022) Physics-informed graph neural Galerkin networks: a unified framework for solving PDE-governed forward and inverse problems. Comput Methods Appl Mech Eng 390:114502MathSciNetCrossRef Gao H, Zahr MJ, Wang J-X (2022) Physics-informed graph neural Galerkin networks: a unified framework for solving PDE-governed forward and inverse problems. Comput Methods Appl Mech Eng 390:114502MathSciNetCrossRef
17.
Zurück zum Zitat Kissas G, Yang Y, Hwuang E, Witschey WR, Detre JA, Perdikaris P (2020) Machine learning in cardiovascular flows modeling: predicting arterial blood pressure from non-invasive 4D flow MRI data using physics-informed neural networks. Comput Methods Appl Mech Eng 358:112623MathSciNetCrossRef Kissas G, Yang Y, Hwuang E, Witschey WR, Detre JA, Perdikaris P (2020) Machine learning in cardiovascular flows modeling: predicting arterial blood pressure from non-invasive 4D flow MRI data using physics-informed neural networks. Comput Methods Appl Mech Eng 358:112623MathSciNetCrossRef
18.
Zurück zum Zitat Lu L, Meng X, Mao Z, Karniadakis GE (2021) Deepxde: a deep learning library for solving differential equations. SIAM Rev 63(1):208–228MathSciNetCrossRef Lu L, Meng X, Mao Z, Karniadakis GE (2021) Deepxde: a deep learning library for solving differential equations. SIAM Rev 63(1):208–228MathSciNetCrossRef
19.
Zurück zum Zitat Glorot X, Bengio Y (2010) Understanding the difficulty of training deep feedforward neural networks. In: Proceedings of the 13th international conference on artificial intelligence and statistics, JMLR workshop and conference proceedings, pp 249–256 Glorot X, Bengio Y (2010) Understanding the difficulty of training deep feedforward neural networks. In: Proceedings of the 13th international conference on artificial intelligence and statistics, JMLR workshop and conference proceedings, pp 249–256
20.
Zurück zum Zitat Zong Y, He Q, Tartakovsky AM (2023) Improved training of physics-informed neural networks for parabolic differential equations with sharply perturbed initial conditions. Comput Methods Appl Mech Eng 414:116125MathSciNetCrossRef Zong Y, He Q, Tartakovsky AM (2023) Improved training of physics-informed neural networks for parabolic differential equations with sharply perturbed initial conditions. Comput Methods Appl Mech Eng 414:116125MathSciNetCrossRef
21.
Zurück zum Zitat Laubscher R, Rousseau P (2022) Application of a mixed variable physics-informed neural network to solve the incompressible steady-state and transient mass, momentum, and energy conservation equations for flow over in-line heated tubes. Appl Soft Comput 114:108050CrossRef Laubscher R, Rousseau P (2022) Application of a mixed variable physics-informed neural network to solve the incompressible steady-state and transient mass, momentum, and energy conservation equations for flow over in-line heated tubes. Appl Soft Comput 114:108050CrossRef
22.
Zurück zum Zitat Laubscher R (2021) Simulation of multi-species flow and heat transfer using physics-informed neural networks. Phys Fluids 33(8):12CrossRef Laubscher R (2021) Simulation of multi-species flow and heat transfer using physics-informed neural networks. Phys Fluids 33(8):12CrossRef
23.
Zurück zum Zitat Li W, Bazant MZ, Zhu J (2021) A physics-guided neural network framework for elastic plates: comparison of governing equations-based and energy-based approaches. Comput Methods Appl Mech Eng 383:113933MathSciNetCrossRef Li W, Bazant MZ, Zhu J (2021) A physics-guided neural network framework for elastic plates: comparison of governing equations-based and energy-based approaches. Comput Methods Appl Mech Eng 383:113933MathSciNetCrossRef
24.
Zurück zum Zitat Yu B et al (2018) The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems. Commun Math Stat 6(1):1–12MathSciNetCrossRef Yu B et al (2018) The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems. Commun Math Stat 6(1):1–12MathSciNetCrossRef
25.
Zurück zum Zitat Luong KA, Le-Duc T, Lee J (2023) Deep reduced-order least-square method—a parallel neural network structure for solving beam problems. Thin-Walled Struct 191:111044CrossRef Luong KA, Le-Duc T, Lee J (2023) Deep reduced-order least-square method—a parallel neural network structure for solving beam problems. Thin-Walled Struct 191:111044CrossRef
26.
Zurück zum Zitat Lu L, Pestourie R, Yao W, Wang Z, Verdugo F, Johnson SG (2021) Physics-informed neural networks with hard constraints for inverse design. SIAM J Sci Comput 43(6):B1105–B1132MathSciNetCrossRef Lu L, Pestourie R, Yao W, Wang Z, Verdugo F, Johnson SG (2021) Physics-informed neural networks with hard constraints for inverse design. SIAM J Sci Comput 43(6):B1105–B1132MathSciNetCrossRef
27.
Zurück zum Zitat Nguyen-Thanh VM, Zhuang X, Rabczuk T (2020) A deep energy method for finite deformation hyperelasticity. Eur J Mech A/Solids 80:103874MathSciNetCrossRef Nguyen-Thanh VM, Zhuang X, Rabczuk T (2020) A deep energy method for finite deformation hyperelasticity. Eur J Mech A/Solids 80:103874MathSciNetCrossRef
28.
Zurück zum Zitat Nguyen-Thanh VM, Anitescu C, Alajlan N, Rabczuk T, Zhuang X (2021) Parametric deep energy approach for elasticity accounting for strain gradient effects. Comput Methods Appl Mech Eng 386:114096MathSciNetCrossRef Nguyen-Thanh VM, Anitescu C, Alajlan N, Rabczuk T, Zhuang X (2021) Parametric deep energy approach for elasticity accounting for strain gradient effects. Comput Methods Appl Mech Eng 386:114096MathSciNetCrossRef
29.
Zurück zum Zitat Liu Z, Yang Y, Cai Q-D (2019) Solving differential equation with constrained multilayer feedforward network. arXiv preprint arXiv:1904.06619 Liu Z, Yang Y, Cai Q-D (2019) Solving differential equation with constrained multilayer feedforward network. arXiv preprint arXiv:​1904.​06619
30.
Zurück zum Zitat Sukumar N, Srivastava A (2022) Exact imposition of boundary conditions with distance functions in physics-informed deep neural networks. Comput Methods Appl Mech Eng 389:114333MathSciNetCrossRef Sukumar N, Srivastava A (2022) Exact imposition of boundary conditions with distance functions in physics-informed deep neural networks. Comput Methods Appl Mech Eng 389:114333MathSciNetCrossRef
31.
Zurück zum Zitat Luong KA, Le-Duc T, Lee J (2023) Automatically imposing boundary conditions for boundary value problems by unified physics-informed neural network. Eng Comput 2:1–23 Luong KA, Le-Duc T, Lee J (2023) Automatically imposing boundary conditions for boundary value problems by unified physics-informed neural network. Eng Comput 2:1–23
32.
Zurück zum Zitat Baydin AG, Pearlmutter BA, Radul AA, Siskind JM (2018) Automatic differentiation in machine learning: a survey. J Mach Learn Res 18:1–43MathSciNet Baydin AG, Pearlmutter BA, Radul AA, Siskind JM (2018) Automatic differentiation in machine learning: a survey. J Mach Learn Res 18:1–43MathSciNet
33.
Zurück zum Zitat Krylov VI, Stroud AH (2006) Approximate calculation of integrals. Courier Corporation, North Chelmsford Krylov VI, Stroud AH (2006) Approximate calculation of integrals. Courier Corporation, North Chelmsford
34.
Zurück zum Zitat Stoer J, Bulirsch R, Bartels R, Gautschi W, Witzgall C (1980) Introduction to numerical analysis, vol 1993. Springer, New YorkCrossRef Stoer J, Bulirsch R, Bartels R, Gautschi W, Witzgall C (1980) Introduction to numerical analysis, vol 1993. Springer, New YorkCrossRef
35.
36.
Zurück zum Zitat Abadi M, Barham P, Chen J, Chen Z, Davis A, Dean J, Devin M, Ghemawat S, Irving G, Isard M et al (2016) Tensorflow: a system for large-scale machine learning. In: Osdi, vol 16. Savannah, pp 265–283 Abadi M, Barham P, Chen J, Chen Z, Davis A, Dean J, Devin M, Ghemawat S, Irving G, Isard M et al (2016) Tensorflow: a system for large-scale machine learning. In: Osdi, vol 16. Savannah, pp 265–283
38.
Zurück zum Zitat Liu DC, Nocedal J (1989) On the limited memory BFGS method for large scale optimization. Math Program 45(1–3):503–528MathSciNetCrossRef Liu DC, Nocedal J (1989) On the limited memory BFGS method for large scale optimization. Math Program 45(1–3):503–528MathSciNetCrossRef
39.
Zurück zum Zitat Chen W, Wang Q, Hesthaven JS, Zhang C (2021) Physics-informed machine learning for reduced-order modeling of nonlinear problems. J Comput Phys 446:110666MathSciNetCrossRef Chen W, Wang Q, Hesthaven JS, Zhang C (2021) Physics-informed machine learning for reduced-order modeling of nonlinear problems. J Comput Phys 446:110666MathSciNetCrossRef
40.
Zurück zum Zitat Timoshenko S, Woinowsky-Krieger S et al (1959) Theory of plates and shells, vol 2. McGraw-Hill, New York Timoshenko S, Woinowsky-Krieger S et al (1959) Theory of plates and shells, vol 2. McGraw-Hill, New York
41.
Zurück zum Zitat Huang Y, Ouyang Z-Y (2020) Exact solution for bending analysis of two-directional functionally graded Timoshenko beams. Arch Appl Mech 90(5):1005–1023CrossRef Huang Y, Ouyang Z-Y (2020) Exact solution for bending analysis of two-directional functionally graded Timoshenko beams. Arch Appl Mech 90(5):1005–1023CrossRef
Metadaten
Titel
A novel normalized reduced-order physics-informed neural network for solving inverse problems
verfasst von
Khang A. Luong
Thang Le-Duc
Seunghye Lee
Jaehong Lee
Publikationsdatum
20.04.2024
Verlag
Springer London
Erschienen in
Engineering with Computers / Ausgabe 5/2024
Print ISSN: 0177-0667
Elektronische ISSN: 1435-5663
DOI
https://doi.org/10.1007/s00366-024-01971-7