hig.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard-cite-them-right
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • sv-SE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • de-DE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
A New Graph-Based Fractality Index to Characterize Complexity of Urban Form
University of Gävle, Faculty of Engineering and Sustainable Development, Department of Computer and Geospatial Sciences, Geospatial Sciences.ORCID iD: 0000-0001-9579-6344
University of Gävle, Faculty of Engineering and Sustainable Development, Department of Computer and Geospatial Sciences, Computer Science. Uppsala universitet.ORCID iD: 0000-0003-0085-5829
University of Gävle, Faculty of Engineering and Sustainable Development, Department of Computer and Geospatial Sciences, Geospatial Sciences.ORCID iD: 0000-0002-3884-3084
University of Gävle, Faculty of Engineering and Sustainable Development, Department of Computer and Geospatial Sciences, Geospatial Sciences. Shenzhen University, China.ORCID iD: 0000-0001-9328-9584
2022 (English)In: ISPRS International Journal of Geo-Information, ISSN 2220-9964, Vol. 11, no 5, article id 287Article in journal (Refereed) Published
Abstract [en]

Examining the complexity of urban form may help to understand human behavior in urban spaces, thereby improving the conditions for sustainable design of future cities. Metrics, such as fractal dimension, ht-index, and cumulative rate of growth (CRG) index have been proposed to measure this complexity. However, as these indicators are statistical rather than spatial, they result in an inability to characterize the spatial complexity of urban forms, such as building footprints. To overcome this problem, this paper proposes a graph-based fractality index (GFI), which is based on a hybrid of fractal theory and deep learning techniques. First, to quantify the spatial complexity, several fractal variants were synthesized to train a deep graph convolutional neural network. Next, building footprints in London were used to test the method, where the results showed that the proposed framework performed better than the traditional indices, i.e., the index is capable of differentiating complex patterns. Another advantage is that it seems to assure that the trained deep learning is objective and not affected by potential biases in empirically selected training datasets Furthermore, the possibility to connect fractal theory and deep learning techniques on complexity issues opens up new possibilities for data-driven GIS science.

Place, publisher, year, edition, pages
MDPI , 2022. Vol. 11, no 5, article id 287
Keywords [en]
complexity; fractals; building groups; graph convolutional neural networks; urban form
National Category
Environmental Sciences Geosciences, Multidisciplinary Cultural Studies
Identifiers
URN: urn:nbn:se:hig:diva-38476DOI: 10.3390/ijgi11050287ISI: 000801418000001Scopus ID: 2-s2.0-85129726341OAI: oai:DiVA.org:hig-38476DiVA, id: diva2:1655069
Available from: 2022-04-29 Created: 2022-04-29 Last updated: 2024-11-26Bibliographically approved
In thesis
1. From Understanding to Generative Design of Sustainable Urban Forms
Open this publication in new window or tab >>From Understanding to Generative Design of Sustainable Urban Forms
2024 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

Urban forms, as a physical proxy of human mobility and activity, are crucial for understanding the mechanisms and factors behind urban dynamics. The rapid expansion of urban sensors and big data has significantly advanced our comprehension of cities and societies. Currently, the development of complexity sciences has led to more sophisticated simulation models, presenting new opportunities for understanding complex urban phenomena. This dissertation integrates fractal geometry theory, deep learning models, and agent-based modelling (ABM) to enhance the understanding and future generative design of sustainable urban forms. Three models were developed in this dissertation based on: Graph-based fractality index (GFI), Spatio-structural self-similarity, Simple agents–complex emergent path systems (SACP), as well as a model and web-based tool for road evaluation by desire path system (RED-PaSS). (1) The GFI model, grounded in fractal theory and deep learning techniques, is capable to characterize the complexity of building groups; (2) the spatio-structural self-similarity model examines self-similarity from a spatial and structural perspective, correcting long-standing misinterpretations in classical statistical fractal theories. The spatio-structural self-similarity model is capable to understand historical urban forms through validation with data from London building groups and US street networks; (3) the SACP model is based on ABM and simulates pedestrian movement based on visibility parameters and simple principles of global destination awareness and local environmental adaptation. The findings of SACP indicate that the angle of vision is crucial for path pattern emergence; and (4) the RED-PaSS model and tool that evaluates road networks by simulating optimal pedestrian paths based on the SACP model. Case studies of 708 US neighbourhood-scale road networks demonstrate RED-PaSS's potential to evaluate, rank, and enhance road networks, improving pedestrian mobility and convenience. This dissertation's holistic approach not only aids in the characterization of current urban patterns but also in generative design of future urban landscapes that are sustainable and resilient. The integration of advanced computational techniques, such as deep learning and ABM, enables exploration of urban dynamics at unprecedented scales and resolutions. The continuous advancement of these models is crucial for addressing urbanization challenges and fostering sustainable, liveable cities.

Abstract [sv]

Stadsformer, som en fysisk proxy för mänsklig mobilitet och aktivitet, är avgörande för att förstå mekanismerna och faktorerna bakom urban dynamik. Hållbarheten i mänskliga aktiviteter fångas och förutsägs också genom studiet av stadsformer. Den snabba expansionen av urbana sensorer och stordata (eng. big data) har avsevärt förbättrat vår förståelse av städer och samhällen. För närvarande har utvecklingen av komplexitetsvetenskap lett till mer sofistikerade modeller, vilket ger nya möjligheter att förstå komplexa urbana fenomen. Denna avhandling integrerar fraktalgeometriteori, djupinlärningsmodeller och agentbaserad modellering (ABM) för att förbättra förståelsen och framtida generativ design av hållbara stadsformer. Fyra modeller utvecklades i denna avhandling baserade på: grafbaserat fraktalitetsindex (GFI), spatio-strukturell självlikhet, enkla agenter–komplexa emergenta vägsystem (SACP), samt en modell med tillhörande webbaserade verktyg för vägutvärdering genom önskat stigsystem (RED-PaSS). (1) GFI-modellen, grundad i fraktalteori och djupinlärningstekniker, kan karakterisera komplexiteten hos byggnadsgrupper; (2) den spatio-strukturella självlikhetsmodellen undersöker självlikhet ur ett rumsligt och strukturellt perspektiv och korrigerar den sedan länge misstolkningen av självlikhet i klassisk statistisk fraktalteori. Denna modell kan också förutsäga årtalen för historiska stadsformer, inklusive byggnadsgrupper i London och gatunät i USA; (3) SACP-modellen är baserad på ABM och simulerar uppkomsten av önskade stigar genom att modellera fotgängares naturliga rörelse med enkla parametrar på syn och interaktionsprinciper; och (4) RED-PaSSmodellen med sitt verktyg utvärderar, rangordnar, och förbättra gångbarheten i vägnät jämfört med de genererade optimala vägsystemen baserade på fotgängares naturliga rörelse. Denna avhandlings holistiska angreppssätt hjälper inte bara till att karakterisera nuvarande urbana mönster utan också i generativ design av framtida urbana landskap som är naturliga och hållbara. Integrationen av avancerade beräkningstekniker, såsom djupinlärning och ABM, möjliggör utforskning av urban dynamik i aldrig tidigare använd skala och upplösning. Den kontinuerliga utvecklingen av dessa modeller är avgörande för att hantera urbaniseringsutmaningar och främja hållbara, levande städer.

Place, publisher, year, edition, pages
Gävle: Gävle University Press, 2024. p. 66
Series
Doctoral thesis ; 52
Keywords
urban forms, agent-based modelling, pedestrian movement, desire paths, fractals, self-similarity, building groups, street networks, generative design, urbanistiska former, agentbaserad modellering, gångrörelser, önskade vägar, fraktaler, självlikhet, byggnadsgrupper, gatunätverk, generativ design
National Category
Geosciences, Multidisciplinary
Identifiers
urn:nbn:se:hig:diva-45835 (URN)978-91-89593-46-6 (ISBN)978-91-89593-47-3 (ISBN)
Public defence
2024-12-17, Lilla Jadwigasalen, Kungsbäcksvägen 47, Gävle, 13:00 (English)
Opponent
Supervisors
Available from: 2024-11-26 Created: 2024-10-14 Last updated: 2024-11-26Bibliographically approved

Open Access in DiVA

fulltext(55410 kB)152 downloads
File information
File name FULLTEXT01.pdfFile size 55410 kBChecksum SHA-512
93dc0f256fd947dcca8c5052c1e5c3697f4be6371d5ecf06ca8effefd6647d3accd858eeb91c912fd9bcbb10090547b0fe305c4390e158ad70ea0326a4dbebf5
Type fulltextMimetype application/pdf

Other links

Publisher's full textScopus

Authority records

Ma, LeiSeipel, StefanBrandt, S. AndersMa, Ding

Search in DiVA

By author/editor
Ma, LeiSeipel, StefanBrandt, S. AndersMa, Ding
By organisation
Geospatial SciencesComputer Science
In the same journal
ISPRS International Journal of Geo-Information
Environmental SciencesGeosciences, MultidisciplinaryCultural Studies

Search outside of DiVA

GoogleGoogle Scholar
Total: 152 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

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 390 hits
CiteExportLink to record
Permanent link

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