kth.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Solving the Schrödinger equation in Higher Dimensions With Physics-Informed Neural Networks
KTH, School of Electrical Engineering and Computer Science (EECS).
2024 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesisAlternative title
Lösning av Schrödingerekvationen i högre dimensioner med fysikinformerade neurala nätverk (Swedish)
Abstract [en]

Physics-Informed Neural Networks (PINNs) are a type of neural network that specializes in solving Partial Differential Equations (PDEs). In this thesis, we implement a PINN to solve the Schrödinger equation in higher spatial dimensions. The Schrödinger equation in higher spatial dimensions is extremely computationally intensive to solve and there have not been many articles focusing on this task. We aim to solve the isotropic harmonic oscillator, the anisotropic harmonic oscillator, and the Woods-Saxon potential. We also perform experiments to evaluate the model’s performance. The research aims to investigate what degree of precision the model can achieve on the potentials. For potentials with analytical solutions, this is answered by calculating the average relative error. The purpose of this study is to develop models and enable new ideas to solve the larger problem known as the many- body problem. The neural network is capable of finding the ground as well as excited states of multiple potentials. However, the effects of the curse of dimensionality persists.

Abstract [sv]

Fysikinformerade neurala nätverk (PINNs) är en typ av neurala nätverk som är specialiserade på att lösa partiella differentialekvationer (PDEs). I denna avhandling implementerar vi en PINN för att lösa Schrödinger-ekvationen i högre rumsdimensioner. Schrödinger-ekvationen i högre rumsdimensioner är extremt beräkningsintensiv att lösa och det har inte funnits många artiklar som fokuserar på denna uppgift. Vi strävar efter att lösa den isotropiska harmoniska oscillatorn, den anisotropiska harmoniska oscillatorn och Woods- Saxon potentialen. Vi utför även experiment för att utvärdera modellens prestanda. Forskningen syftar till att undersöka vilken grad av precision modellen kan uppnå på potentialerna. För potentialer med analytiska lösningar kommer detta att besvaras genom att beräkna det genomsnittliga relativa felet. Syftet med denna studie är att utveckla modeller och möjliggöra nya idéer för att lösa det större problemet som kallas mångakroppsproblemet. Det neurala nätverket kan hitta grund såväl som exciterade tillstånd för flera potentialer. Däremot kvarstår effekten av dimensionalitetens förbannelse.

Place, publisher, year, edition, pages
2024. , p. 47
Series
TRITA-EECS-EX ; 2024:893
Keywords [en]
Quantum Mechanics, Schrödinger Equation, Machine Learning, Physics- Informed Neural Network, Higher Dimensions
Keywords [sv]
Kvantmekanik, Schrödingerekvation, Maskininlärning, Fysikinformerade neu- rala nätverk, Högre dimensioner
National Category
Computer and Information Sciences
Identifiers
URN: urn:nbn:se:kth:diva-360861OAI: oai:DiVA.org:kth-360861DiVA, id: diva2:1942242
Supervisors
Examiners
Available from: 2025-03-11 Created: 2025-03-04 Last updated: 2025-03-11Bibliographically approved

Open Access in DiVA

fulltext(3476 kB)171 downloads
File information
File name FULLTEXT01.pdfFile size 3476 kBChecksum SHA-512
2ccb22112d23cbb94a99de4ff2a28693e13471b06cfa39058876ed842a477b5e8d82dde0dc319cb1932a7777bb5602c6ae8b8a5ca3ed255c905de98c8c78a9f5
Type fulltextMimetype application/pdf

By organisation
School of Electrical Engineering and Computer Science (EECS)
Computer and Information Sciences

Search outside of DiVA

GoogleGoogle Scholar
Total: 171 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 349 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf