Scholarlink

<p>Accurately quantifying urban renewal potential (URP) and understanding urban landscape influences on URP are critical for identifying renewal areas and improving intervention effectiveness. However, existing studies still lack a systematic and intelligent URP assessment framework, and the underlying relationships between landscape elements and URP remain insufficiently understood. Therefore, this study proposes a six-dimensional URP assessment framework that integrates large language models (LLMs) and the Genetic Algorithm–Particle Swarm Optimisation (GAPSO) algorithm for intelligent measurement, and applies interpretable machine learning to uncover nonlinear influences and threshold effects. Results show that high-potential blocks in Wuhan concentrate in waterfront areas of the urban core, with URP peaking around 2&#xa0;km from the city centre. Overall, mean building area, building expandability, and public transportation convenience are the three most influential elements, contributing 23.77%, 11.21%, and 5.37%, respectively. All three exhibit U-shaped relationships with URP, reversing from negative to positive beyond 102.3, 3.9, and 31.2, respectively. The study further identifies strong bivariate interaction effects among the landscape elements. For example, low values of mean building area and building expandability interact negatively, whereas high values of public transportation convenience and road intersection quantities show positive synergy. Moreover, landscape elements’ influence mechanisms exhibit marked typological heterogeneity across high-potential block types (comprehensive enhancement, high economic return, and perception enhancement), with divergent dominant factors, shifting thresholds, and varying interactions. By applying artificial intelligence to complex renewal scenarios, this study provides a replicable, data-driven framework and empirical evidence for planners to identify street-block renewal priorities and implement precision-targeted interventions.</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

Intelligent Assessment of Urban Renewal Potential and the Nonlinear Relationships with Landscape Elements: Integrating Large Language Models with Interpretable Machine Learning

  • 期刊论文

  • 发表日期：

    2026年7月20日

查看全文
  • Yukun Jiang,
  • Junhong Liang,
  • Qingsong He

摘要

Accurately quantifying urban renewal potential (URP) and understanding urban landscape influences on URP are critical for identifying renewal areas and improving intervention effectiveness. However, existing studies still lack a systematic and intelligent URP assessment framework, and the underlying relationships between landscape elements and URP remain insufficiently understood. Therefore, this study proposes a six-dimensional URP assessment framework that integrates large language models (LLMs) and the Genetic Algorithm–Particle Swarm Optimisation (GAPSO) algorithm for intelligent measurement, and applies interpretable machine learning to uncover nonlinear influences and threshold effects. Results show that high-potential blocks in Wuhan concentrate in waterfront areas of the urban core, with URP peaking around 2 km from the city centre. Overall, mean building area, building expandability, and public transportation convenience are the three most influential elements, contributing 23.77%, 11.21%, and 5.37%, respectively. All three exhibit U-shaped relationships with URP, reversing from negative to positive beyond 102.3, 3.9, and 31.2, respectively. The study further identifies strong bivariate interaction effects among the landscape elements. For example, low values of mean building area and building expandability interact negatively, whereas high values of public transportation convenience and road intersection quantities show positive synergy. Moreover, landscape elements’ influence mechanisms exhibit marked typological heterogeneity across high-potential block types (comprehensive enhancement, high economic return, and perception enhancement), with divergent dominant factors, shifting thresholds, and varying interactions. By applying artificial intelligence to complex renewal scenarios, this study provides a replicable, data-driven framework and empirical evidence for planners to identify street-block renewal priorities and implement precision-targeted interventions.