Journal of Transportation Systems Engineering and Information Technology ›› 2019, Vol. 19 ›› Issue (4): 28-32.

• Forum about Comprehensive Transportation System • Previous Articles     Next Articles

Capturing Car Ownership Behavior Considering Spatial Autocorrelation in Traffic Analysis Zones

WANG Xiao-quan, SHAO Chun-fu, YIN Chao-ying, DONG Chun-jiao   

  1. Key Laboratory of Transport Industry of Big Data Application Technologies for Comprehensive Transport, Beijing Jiaotong University, Beijing 100044, China
  • Received:2019-01-22 Revised:2019-04-14 Online:2019-08-25 Published:2019-08-26

考虑交通小区相关性的小汽车拥有行为研究

王晓全,邵春福*,尹超英,董春娇   

  1. 北京交通大学综合交通运输大数据应用技术交通运输行业重点实验室,北京 100044
  • 作者简介:王晓全(1992-),男,黑龙江宾县人,博士生.
  • 基金资助:

    中央高校基本科研业务经费专项资金/Fundamental Research Funds for the Central Universities of Ministry of Education of China(2019YJS101,2017JBZ106).

Abstract:

In order to investigate the influence of built environment factors on household car ownership, a multilevel Bayesian model is employed considering the spatial autocorrelation among traffic analysis zones (TAZs). In the Bayesian model, three types of adjacent matrix are used to specify the spatial autocorrelation term including 0-1 adjacent matrix, common boundary adjacent matrix and centroid distance adjacent matrix. The models are calibrated based on the Changchun household travel survey data. The result shows that the spatial autocorrelation exists significantly. Among the calibrated models, the multilevel Bayesian model with common boundary adjacent matrix fit the data best. Additionally, residential density, land use mix, intersection density and transit station density all have significantly negative influence on household car ownership. It suggests that it can be effective for reducing car ownership by optimizing urban built environment.

Key words: traffic engineering, built environment, car ownership, spatial autocorrelation, multilevel Bayesian model

摘要:

为分析建成环境对家庭小汽车拥有的影响,考虑交通小区间的空间相关性,分别基于0-1 邻接矩阵、共同边界邻接矩阵及质心空间距离矩阵构建了层次Bayesian 模型,并与不考虑空间相关性的模型结果进行了对比.基于长春居民出行调查数据对模型参数进行估计,结果表明:交通小区间的空间相关性显著存在;以不考虑空间相关性的模型作为对比,基于公共边界邻接矩阵的层次Bayesian 模型拟合效果最优;在控制家庭层面社会经济变量后,居住密度、土地利用混合度、交叉口密度及公共交通站点密度均对家庭小汽车拥有具有显著的负向效应,表明通过优化城市建成环境策略可有效抑制小汽车拥有量的增长.

关键词: 交通工程, 建成环境, 小汽车拥有, 空间自相关, 层次Bayesian模型

CLC Number: