网约车对公共交通的潜在替代效应测度及影响因素——以成都市为例  

Measuring the potential substitution effect of ride-hailing travel on public transport and its influencing factors:A case study of Chengdu

作　　者：郑智成 张丽君[1,2] 秦耀辰[1,2] 荣培君 李阳[1] 张晶飞 ZHENG Zhicheng;ZHANG Lijun;QIN Yaochen;RONG Peijun;LI Yang;ZHANG Jingfei(College of Geographical Sciences/Key Laboratory of Geospatial Technology for the Middle and Lower Yellow River Regions,Ministry of Education,Faculty of Geographical Science and Engineering,Henan University,Zhengzhou 450046,China;Key Research Institute of Yellow River Civilization and Sustainable Development&Collaborative Innovation Center on Yellow River Civilization Jointly Built by Henan Province and Ministry of Education,Henan University,Kaifeng 475004,Henan,China;College of Collaborative Innovation Center on Urban and Rural Harmonious Development of Henan Province,Henan University of Economics and Law,Zhengzhou 450046,China)

机构地区：[1]河南大学地理科学与工程学部地理科学学院/黄河中下游数字地理技术教育部重点实验室,郑州450046 [2]河南大学黄河文明与可持续发展研究中心暨黄河文明省部共建协同创新中心,开封475004 [3]河南财经政法大学城乡协调发展河南省协同创新中心,郑州450046

出　　处：《地理学报》2025年第2期503-522,共20页Acta Geographica Sinica

基　　金：国家自然科学基金项目(42171295,42071294,42101206);黄河文明省部共建协同创新中心重大项目(2020M19)。

摘　　要：数字交通背景下,网约车与公共交通使用和建成环境的关系是城市地理学和交通地理学共同关注的重点论题。然而,目前仍缺乏从时空维度量化网约车对公共交通出行的影响,也未明确建成环境在其中发挥的作用。通过引入需求弹性理论与大数据技术,提出网约车出行替代效应时空测度方法,并发展了一种高效、动态、精细化的建成环境量化方法。运用随机森林和可解释机器学习模型,重点剖析了影响替代效应的多因素非线性交互与时空耦合作用。成都市的实证研究表明:(1)出行效率是影响公共交通竞争力的关键。同一行程下的公共交通行程时间是网约车的2.0~3.5倍,其中需花费10 min的步行时间来完成最初/最后1 km,且需要0~2次换乘。(2)中心城区的网约车出行对公共交通产生明显替代效应,工作日和休息日分别有28.69%和27.08%的网约车出行替代了公共交通,并在高峰时段明显增强。(3)目的地可达性对替代率的正向贡献度最高,其次为人口社会经济要素,城市空间形态和公共交通可达性对替代率的影响程度相对较小。(4)建成环境对替代效应的影响呈现出非线性交互特征,且阈值区间和交互影响强度随时间的动态变化而差异显著。研究不仅突破了传统静态“空间—行为”研究局限性,也为城市交通出行优化和建成环境的精细化调控提供了应用参考。In the context of digital transportation,the relationship between ride-hailing services,public transport usage,and the built environment is a crucial area of research in urban geography and transportation geography.Nevertheless,there remains a lack of research in quantifying the impact of ride-hailing on public transport travel from the spatio-temporal dimensions,and the role played by the built environment in this context has not been clarified.To address these limitations,a spatio-temporal measure method of the substitution effect of ride-hailing travel was proposed by introducing demand elasticity theory and big data technology.Additionally,an efficient,dynamic,and refined measure method of the built environment was developed.On this basis,by integrating random forests with interpretable machine learning models,this paper focused on analyzing the multi-factorial nonlinear interactions and spatio-temporal coupling effect that influence the substitution effect.The empirical study of Chengdu shows that:(1)Travel efficiency is key to the competitiveness of public transport.For the same trip,the travel time of public transport is typically 2.0-3.5 times that of ride-hailing.Specifically,taking public transport often requires a 10-min walk to complete the first/last 1 km and involves 0-2 transfers.(2)There is a significant substitution effect of ride-hailing travel on public transport in the central city,with 28.69%and 27.08%of ride-hailing trips substituting public transport on weekdays and weekends,respectively,and the substitution effect is significantly enhanced during peak periods.(3)Destination accessibility has the highest positive impact on the substitution rate,followed by demographic socioeconomic factors,with urban spatial form and public transport accessibility having a relatively small degree of influence on the substitution rate.(4)The influence of built environment on the substitution effect exhibits a nonlinear interactive characteristic,with threshold ranges and interaction strengths showing signifi

关 键 词：网约车 替代效应 城市建成环境 动态人地关系 非线性 阈值效应 成都市 

分 类 号：F572.88[经济管理—产业经济] F724.6