摘要
利用网约车订单数据,提出基于网络距离搜索的核密度估计方法进行载客热点识别,分析网约车载客热点的空间分布特性。建立以兴趣点(POI)、土地利用多样性、路网密度以及公共交通临近性作为解释变量、载客点网络核密度值作为因变量的地理加权回归模型(GWR),分析各类解释变量回归系数分布的空间异质性,挖掘了城市建成环境在空间区域上对网约车出行的不同影响。研究结果表明:从宏观角度,餐饮住宅、风景医疗、科教购物设施、土地利用多样性与载客点网络核密度值总体呈正相关性;从微观角度,各类设施对载客点网络核密度值的影响程度在不同区域存在明显的时空异质性,且异质性是受到多种因素共同作用的结果。土地利用多样性、道路属性和交通可达性的影响作用在时间上具有一致性。
In order to analyze the relationship between passengers’travel demand of online car-hailing and urban built environment,this paper proposes a network kernel density estimation method based on network distance search to identify the pick-up hot spots of online car-hailing.The network kernel density estimation considers the difference of travelers’pick-up behavior between the road sections and intersections,which reflects the spatial distribution heterogeneity of online car-hailing demand.The regression model of geographically weighted regression model(GWR)is set up,taking the amount of points of interest(POI),land use diversity,road network density,public transport proximity as explanatory variables and passenger pick-up points’network kernel density as dependent variable.Spatial distribution heterogeneity of all explanatory variables’regression coefficient has been analyzed,and then excavate the different impact of urban built environment on online car-hailing demand.The results show that,catering-housing facilities,scenic-medical treatment facilities,education-shopping facilities,and land-use diversity are generally positively correlated with the pick-up spots network kernel density values from the macroscopic perspective.Further,the various facilities’influence on the network kernel density value of passenger pick-up points shows obvious spatiotemporal heterogeneity in different regions from a microscopic point perspective.The spatiotemporal heterogeneity is affected by various factors together.However,the influence of three kinds of explanatory variables on pick-up hot spots shows temporal consistent,including land use diversity,road attributes and public transportation accessibility.The research results can be applied for urban planning and improvision.Optimizing the urban built environment distribution can effectively promote passengers’travel demand to enhance the urban vitality.At the same time,the results provide theory basis for managers to relieve traffic congestion by changing POI facilities’distribution and improving public transport’s convenience.
作者
龙雪琴
赵欢
周萌
毛健旭
陈亦新
Long Xueqin;Zhao Huan;Zhou Meng;Mao Jianxu;Chen Yixin(College of Transportation Engineering,Chang'an University,Xi'an 710064,Shaanxi,China)
出处
《地理科学》
CSSCI
CSCD
北大核心
2022年第12期2076-2084,共9页
Scientia Geographica Sinica
基金
国家重点研发计划项目(2019YBFB1600500)
陕西省自然科学基础研究计划(2021JQ-276)资助。
关键词
出行行为
载客热点
网络核密度
地理加权回归模型
时空分异性
travel behavior
passengers’pick-up hot spots
network kernel density
geographically weighted regression model
spatiotemporal heterogeneity
