Scholarlink

<p>Urbanization profoundly affects city dynamics worldwide, creating a pressing need for effective spatial zoning to balance urban development with sustainability. This study introduces a novel spatial zoning framework for a district by integrating urban growth simulation with a dynamic, multidimensional index system at the township level, aiming to enhance the precision and sustainability of urban planning. The methodology employs the MDBN-CA model, a hybrid approach combining Cellular Automata for spatial dynamics, Deep Belief Networks to capture complex urban patterns, and the Modified Bat Algorithm for optimization. Urban expansion is simulated through 2035, while township-level indicators encompassing morphology, socioeconomic factors, ecological impacts, and growth trends inform an evaluation system. K-means clustering is applied to classify the district into six distinct zones for 2035: Ecological Priority, Potential, Radiation, Optimize, Highly Urbanized, and Key Urbanized zones. Results demonstrate that the MDBN-CA model outperforms the ANN-CA model, achieving improved simulation accuracy (0.727) and spatial pattern coherence (86.05%). The district is projected to undergo significant urban expansion by 2035, presenting both advantages and challenges such as inefficient land use and uneven development. The study concludes by recommending strategies to optimize land use, enhance urban functions, restore spatial patterns, and promote sustainable, eco-friendly development, offering valuable guidance for future urban planning efforts.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A hybrid MDBN-CA framework for simulating urban growth and optimizing multidimensional zoning in sustainable urban planning

  • 期刊论文

  • 发表日期：

    2026年5月14日

查看全文
  • Esha Malik,
  • Muhammad Nadeem,
  • Jizong Jiao,
  • Mohammad Arsalan Khan,
  • Abdullah Faiz Al Asmari

摘要

Urbanization profoundly affects city dynamics worldwide, creating a pressing need for effective spatial zoning to balance urban development with sustainability. This study introduces a novel spatial zoning framework for a district by integrating urban growth simulation with a dynamic, multidimensional index system at the township level, aiming to enhance the precision and sustainability of urban planning. The methodology employs the MDBN-CA model, a hybrid approach combining Cellular Automata for spatial dynamics, Deep Belief Networks to capture complex urban patterns, and the Modified Bat Algorithm for optimization. Urban expansion is simulated through 2035, while township-level indicators encompassing morphology, socioeconomic factors, ecological impacts, and growth trends inform an evaluation system. K-means clustering is applied to classify the district into six distinct zones for 2035: Ecological Priority, Potential, Radiation, Optimize, Highly Urbanized, and Key Urbanized zones. Results demonstrate that the MDBN-CA model outperforms the ANN-CA model, achieving improved simulation accuracy (0.727) and spatial pattern coherence (86.05%). The district is projected to undergo significant urban expansion by 2035, presenting both advantages and challenges such as inefficient land use and uneven development. The study concludes by recommending strategies to optimize land use, enhance urban functions, restore spatial patterns, and promote sustainable, eco-friendly development, offering valuable guidance for future urban planning efforts.