hh.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
Machine Learning Cascade for Defect Detection in Graphene
Halmstad University, School of Information Technology.
Halmstad University, School of Information Technology.
2026 (English)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesis
Abstract [en]

Graphene is a two-dimensional material with exceptional mechanical, electrical, and thermal properties, but its performance is highly sensitive to structural defects that arise during production. This thesis investigated whether a machine learning model could determine the type, orientation, and position of structural defects in graphene from a small number of topological probe measurements, without directly observing the defect. Graphene was modelled as a graph, where carbon atoms are represented as nodes and bonds as edges, and four defect types were considered: single vacancy, double vacancy, Stone-Wales, and 555-777. The measurement signal exploited the fact that the perfect graphene lattice contains no odd-length cycles, while defects introduce odd-length cycles locally whose length grows with distance from the defect. A synthetic dataset of 100,000 samples was generated using an adaptive probing strategy that placed six probes per sample, guided by breadth-first search to estimate the defect location. A three-stage cascade of machine learning models was then trained to predict defect type, orientation, and position sequentially. The cascade achieved a defect type accuracy of 96.4%, an average orientation accuracy of 88.2%, and a mean localisation distance of 0.283 lattice nodes, outperforming a flat baseline classifier by a factor of 4.5 in localisation accuracy. The results show that topological measurements alone carry sufficient information for accurate defect classification and localisation in graphene.

Abstract [sv]

Grafén är ett tvådimensionellt material med exceptionella mekaniska, elektriska och termiska egenskaper, men dess prestanda är mycket känslig för strukturella defekter som uppstår under produktion. Denna avhandling undersökte om en maskininlärningsmodell kan bestämma typ, orientering och position för strukturella defekter i grafen utifrån ett litet antal topologiska probmätningar, utan att direkt observera defekten. Grafén modellerades som en graf, där kolatomer representeras som noder och bindningar som kanter, och fyra defekttyper undersöktes: enkelvakans, dubbelvakans, Stone-Wales och 555-777. Mätsignalen utnyttjade det faktum att perfekt grafén inte innehåller några udda cykler, medan defekter introducerar udda cykler lokalt vars längd ökar med avståndet från defekten. Ett syntetiskt dataset med 100 000 prover genererades med hjälp av en adaptiv probstrategi som placerade sex prober per prov, vägledd av Breadth-First Search för att uppskatta defektens position. En trestegskaskad av maskininlärningsmodeller tränades sedan för att sekventiellt förutsäga defekttyp, orientering och position. Kaskaden uppnådde en defekttypsnoggrannhet på 96,4%, en genomsnittlig orienteringsnoggrannhet på 88,2% och ett genomsnittligt lokaliseringsavstånd på 0,283 noder, vilket överträffade en platt bas-klassificerare med en faktor på 4,5 i lokaliseringsnoggrannhet. Resultaten visar att topologiska mätningar ensamma innehåller tillräcklig information för noggrann defektklassificering och lokalisering i grafén.

Place, publisher, year, edition, pages
2026. , p. 45
Keywords [en]
graphene, defect detection, machine learning, graph theory, topological measurements, odd-cycle distance, cascade classifier, synthetic data, breadth-first search
Keywords [sv]
grafén, defektdetektering, maskininlärning, grafteori, topologiska mätningar, udda cykelavstånd, kaskadklassificerare, syntetiska data
National Category
Artificial Intelligence
Identifiers
URN: urn:nbn:se:hh:diva-59407OAI: oai:DiVA.org:hh-59407DiVA, id: diva2:2071603
Supervisors
Examiners
Available from: 2026-06-15 Created: 2026-06-14 Last updated: 2026-06-15Bibliographically approved

Open Access in DiVA

fulltext(2254 kB)64 downloads
File information
File name FULLTEXT02.pdfFile size 2254 kBChecksum SHA-512
8d6a561b520f586f2143c21f694076c2f73af85ff39f637570940db635a4674b8ddd46eb2a4c3b1cfd140691066cdcf87d4b7d837e4578a38416bfeb271e3b4a
Type fulltextMimetype application/pdf

By organisation
School of Information Technology
Artificial Intelligence

Search outside of DiVA

GoogleGoogle Scholar
Total: 64 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: 202 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