期刊文献+

建成环境对常规公交空间公平性的非线性影响

Nonlinear Influence of Built Environment on Spatial Fairness of Conventional Transit
原文传递
导出
摘要 为揭示显著影响常规公交空间公平性的建成环境因子及其有效范围,建立了交通分析区尺度的常规公交空间公平性测度模型(空间排斥指数模型)。融合常规公交网络和城市POI等多源数据,通过Voronoi图、网络分析等方法测算了昆明市主城5区949个交通分析区的建成环境3D(密度、距离、多样性)因子和基于可达性的公交空间排斥指数。使用梯度提升树(GBDT)模型拟合了建成环境因子与常规公交空间公平性之间的复杂关系,并与多元回归模型结果进行了对比。基于变量重要度方法和偏依赖图揭示了建成环境因子对常规公交空间公平性的影响。结果表明:SEI模型可以反映每个交通分析区的交通公平情况,能够在研究区域上辨识交通可达性的劣势范围;与经典的线性回归模型相比,非线性的GBDT模型具有更高拟合优度和解释力;建成环境与公交空间排斥指数之间具有复杂的非线性关系,建成环境在影响常规公交空间公平性的过程中呈现出明显的阈值效应;“距离”指标(44.01%)和“密度”指标(43.89%)的影响大于“多样性”指标(12.08%),行政中心邻近度(31.567%)、居住密度(18.917%)和公交站点密度(13.846%)是建成环境层面影响交通公平的关键要素。研究结果可为基于土地利用和空间规划的交通公平改善策略制订提供参考。 To reveal the built environmental factors that significantly affect the spatial fairness of conventional transit and their effective ranges,a spatial fairness measurement model(spatial exclusion index model)for conventional transit at the scale of traffic analysis zone is established.Using conventional transit network data and urban POI data,the built environment 3D(density,distance,diversity)factors and the transit spatial exclusion index based on reachability of 949 traffic analysis areas in 5 main urban districts of Kunming are calculated by using Voronoi diagram,network analysis and other methods.The complex relationship between built environment factors and the spatial fairness of conventional transit is fitted by using gradient lifting tree(GBDT)model,and the result is compared with the result of multiple regression model.The influence of built environmental factors on the spatial fairness of conventional transit is revealed based on the variable importance method and partial dependence graph.The result shows that(1)SEI model can reflect the traffic fairness of each traffic analysis area,and can identify the disadvantage range of traffic reachability in the study area;(2)compared with the classical linear regression model,the nonlinear GBDT model has a higher goodness of fit and explanatory power;(3)there is a complex nonlinear relationship between the built environment and the transit spatial exclusion index,and the built environment shows an obvious threshold effect in the process of affecting the spatial fairness of conventional transit;(4)the influences of“distance”(44.01%)and“density”(43.89%)are greater than that of the“diversity”(12.08%),and administration center proximity(31.57%),residential density(18.98%)and transit stop density(13.85%)are the key factors that affect traffic fairness at the level of built environment.The research result can provide reference for the formulation of traffic fairness improvement strategy based on land use and spatial planning.
作者 李武 戢晓峰 陈方 刘传颖 LI Wu;JI Xiao-feng;CHEN Fang;LIU Chuan-ying(School of Transportation,Dalian University of Technology,Dalian Liaoning 116024,China;Yunnan Integrated Transport Development and Regional Logistics Management Think Tank,Kunming Yunnan 650504,China;School of Traffic Engineering,Kunming University of Science and Technology,Kunming Yunnan 650504,China;School of Marxism,Kunming University of Science and Technology,Kunming Yunnan 650504,China;School of Transportation Engineering,Dalian Maritime University,Dalian Liaoning 116026,China)
出处 《公路交通科技》 CSCD 北大核心 2023年第8期207-213,共7页 Journal of Highway and Transportation Research and Development
基金 国家自然科学基金项目(71904068,41501174) 云南省创新引导与科技型企业培育计划项目(202004AR040022)。
关键词 城市交通 影响分析 可解释机器学习算法 GBDT模型 交通公平 建成环境 urban traffic influence analysis interpretable machine learning algorithm GBDT model traffic fairness built environment
  • 相关文献

参考文献6

二级参考文献71

共引文献76

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部