建成环境对步行出行强度的影响分析——以南京市中心城区为例  

Analysis of the Influence of Built Environment on Walking Intensity——A Case Study of Central Urban Area of Nanjing

作　　者：余忠洛 费跃 温旭丽[1] 张强 YU Zhongluo;FEI Yue;WEN Xuli;ZHANG Qiang(School of Civil and Traffic Engineering,Chengxian College,Southeast University,Nanjing 210088,China;China International Engineering Consulting Corporation,Beijing 100048,China)

机构地区：[1]东南大学成贤学院土木与交通工程学院,南京210088 [2]中国国际工程咨询有限公司,北京100048

出　　处：《交通工程》2025年第5期38-44,73,共8页Journal of Transportation Engineering

基　　金：江苏省高等学校自然科学研究基金面上项目(21KJB580019,21KJD580004);东南大学成贤学院国家级科研项目培育基金(2023NCF001)。

摘　　要：以南京市中心城区为研究范围,在划分步行交通分析小区的基础上,结合POI数据,路网数据、手机信令数据等多源数据,计算分析小区内的建成环境指标和步行出行强度,以建成环境指标为自变量,步行出行强度为因变量,利用梯度提升决策树模型进行回归分析,基于Shapley additive explanations(SHAP)对变量影响进行分析。结果表明,就业岗位密度、用地混合熵指数和到市中心的距离对步行出行强度预测值的贡献较大,且具有一定的阈值效应和空间分布差异性。Takes the central urban area of Nanjing as the research scope,divides the walking traffic analysis districts,calculates the built environment index and walking trip intensity by combining the multi-source data such as POI data,road network data and mobile signaling data in the walking analysis districts.With the built environment index as the independent variable and walking trip intensity as the dependent variable,regression analysis was carried out by using the Gradient Boosting Decision Tree model,and Shapley additive explanations(SHAP)were used to analysis the influence of variables.The results show that job density,land use mixed entropy index and distance to the city center contributed significantly to the predicted value of walking intensity,with a certain threshold effect and spatial distribution differences.

关 键 词：建成环境 步行出行强度 梯度提升决策树 SHAP 空间分布 

分 类 号：U125[交通运输工程]