Scholarlink

<p>Understanding traffic energy consumption at the county level is crucial for advancing low-carbon urban development, especially under China’s dual carbon goals. Yet, limited data and model applicability have hindered county-level research. This study introduces a prediction model integrating urban morphology and gas station sales data, using Lianjiang County—a typical monocentric Chinese county—as a case. Data from gas station sales, field surveys, and spatial analysis informed energy estimates and urban morphology indicators (e.g., road density, POI (Point of Interest) density, road hierarchy). A BP neural network model was developed to analyze spatial patterns and driving factors of energy consumption. The results show that traffic energy consumption exhibits clear spatial heterogeneity across neighborhoods. Areas with higher road density and mixed road hierarchies tend to have higher consumption, while areas with better bus accessibility and walkability show lower energy use. POI density was found to indirectly influence consumption by shaping travel behavior. The BP neural network model achieved a high R² value (<b>&gt; 0.8</b>) and low MSE, confirming its predictive capability under data-limited conditions. This study contributes theoretical and methodological insights for energy-efficient transportation planning in small and medium-sized cities and offers practical guidance for policy design aligned with China’s carbon reduction targets.</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

How do urban morphology and gas station data reflect transportation energy consumption?

  • 期刊论文

  • 发表日期：

    2025年12月22日

查看全文
  • Xin-Chen Hong,
  • Jingsong Lin,
  • Yiyang Wang,
  • Shiying Li,
  • Huagui Guo

摘要

Understanding traffic energy consumption at the county level is crucial for advancing low-carbon urban development, especially under China’s dual carbon goals. Yet, limited data and model applicability have hindered county-level research. This study introduces a prediction model integrating urban morphology and gas station sales data, using Lianjiang County—a typical monocentric Chinese county—as a case. Data from gas station sales, field surveys, and spatial analysis informed energy estimates and urban morphology indicators (e.g., road density, POI (Point of Interest) density, road hierarchy). A BP neural network model was developed to analyze spatial patterns and driving factors of energy consumption. The results show that traffic energy consumption exhibits clear spatial heterogeneity across neighborhoods. Areas with higher road density and mixed road hierarchies tend to have higher consumption, while areas with better bus accessibility and walkability show lower energy use. POI density was found to indirectly influence consumption by shaping travel behavior. The BP neural network model achieved a high R² value (> 0.8) and low MSE, confirming its predictive capability under data-limited conditions. This study contributes theoretical and methodological insights for energy-efficient transportation planning in small and medium-sized cities and offers practical guidance for policy design aligned with China’s carbon reduction targets.