交通运输工程与信息学报2024，Vol.22Issue(4)：67-84,18.DOI:10.19961/j.cnki.1672-4747.2024.05.011
考虑时空自相关性的共享电动汽车出行选择影响因素分析
Analysis of factors influencing shared electric vehicle travel choice considering spatiotemporal autocorrelation
摘要
Abstract
As an emerging low-carbon mode of transportation,shared electric vehicles(SEVs)have the potential to optimize traffic resource allocation and promote the sustainable development of ur-ban transportation systems.Therefore,it is critical to study their spatiotemporal distribution patterns and influencing factors.This study investigated the spatiotemporal autocorrelation characteristics of SEV trips.It then integrated an improved regional aggregation method with a refined definition of Points of Interest(POIs)spatial distribution patterns,which include density,proximity,and agglomer-ation.A comprehensive"S-T+S+T+5D"quantification system for influencing factors was estab-lished,which considered spatiotemporal features,station attributes,and built environment character-istics.Finally,by constructing and comparing multiple identification models,this study revealed the extent and mechanisms of various influencing factors.The results indicate the following:(1)Com-pared to the Generalized Linear Model(GLM)and Random Forest(RF)models,the Generalized Ad-ditive Mixed Model(GAMM),which considers geographical interaction effects,can more accurately identify influencing factors and effectively explain the spatial dependency of service demand;(2)Among the influencing factors,station attributes have the most significant impact;parking capacity exhibits a threshold effect,where an area capacity exceeding 70 units can suppress travel demand,and overly close proximity between stations can intensify internal competition;(3)SEVs have the po-tential to compensate for insufficient urban public transportation services,particularly in areas with weaker service coverage more than two kilometers from metro entrances and around major passen-ger transport hubs;(4)The intensity of land-use mix exhibits an inverted U-shaped nonlinear relation-ship,and the density of entertainment venues,proximity to universities,and agglomeration of medi-cal facilities all promote SEV travel.These findings provide theoretical support for operators when conducting short-term station optimization and long-term layout planning.关键词
城市交通/非线性分析/广义可加混合模型/共享电动汽车/时空自相关Key words
urban traffic/nonlinear analysis/generalized additive mixed model/shared electric vehi-cles/spatiotemporal autocorrelation分类
交通工程引用本文复制引用
廖洋,罗霞,王红杰..考虑时空自相关性的共享电动汽车出行选择影响因素分析[J].交通运输工程与信息学报,2024,22(4):67-84,18.基金项目
四川省科技计划项目(2024NSFSC0179) （2024NSFSC0179）
