With the expansion of urban space and the rapid growth of car ownership in China,urban traffic congestion and environmental pollution have become more serious.In the face of this problem,the transportation planning department proposed to vigorously develop urban rail transit.On the one hand,it hopes that rail transit will connect highdensity residential areas and employment centers,allowing more commuters to choose the way of rail;on the other hand,it hopes to improve public transport Accessibility,guiding residents to travel from cars to public transportation,thereby alleviating urban congestion.Therefore,more and more scholars began to pay attention to the relationship between rail transit and car travel,and tried to use research results to guide transportation planning and management has become a current research hotspot.However,in reality,the replacement effect of rail transit on cars has not been as expected,and the traffic jam in the central urban area with higher density of rail and bus line networks is the most serious.The reason for this dilemma is largely due to the fact that when planners allocate space facilities,they mostly only focus on the macroscale or static distribution of urban occupational housing space,while ignoring the spatial characteristics and dynamics of micro-behavior based on travellers.Commuter network characteristics.In fact,the choice of commuting methods by residents,in addition to working distance,more consideration is the actual accessibility of both ends of the work,especially for rail transit users,the more important thing is whether the track covers the space at both ends of the work and living range.In addition,due to the extremely uneven development of urban internal space,coupled with the obvious spatial differences in the location of rail transit stations,this spatial heterogeneity will further promote social differentiation through market behaviors such as housing prices and promote rail transit There are differences in the impact on residents’ commuting.Therefore,the impact of rail transit on car commuting should pay more attention to the micro-behavior and micro-socioeconomic attributes of travellers.This article adopts a macro-micro nested analysis perspective.First,28936 total commuting samples are selected from the 2016 travel survey data of residents in Kunming.From the macro level,the urban living space,working space and urban living space are analyzed based on the GIS kernel density and OD matrix space method.Commute flow distribution characteristics,and use the GIS bivariate Moran index to measure the matching relationship between rail transit accessibility and car travel intensity to evaluate the current status effect between rail transit and travel behavior;secondly,turn to the micro behavior level,according to the behavior of residents The coverage relationship between the space and the rail station is divided into three types of residents along the track,and the spatial and statistical methods are used to analyze the characteristics of the residents’ commuting behavior along the line,in order to assess the potential impact capacity and effect of rail transit on the residents’ commuting behavior;finally,according to the track Six types of cluster samples were extracted from the spatial heterogeneity of the station.Based on the selection of key influencing factors,a binary logistic model influencing the travel mode of cars was constructed to verify and analyze the impact mechanism of rail transit on the travel of cars by residents.Research results:(1)The overall spatial distribution of occupation and residence of the total commuting sample is basically consistent with the current urban land use distribution of Kunming City.Some medium and high-density residential areas and employment areas are concentrated along the track.For stretching the urban spatial framework,Facilitating important urban transport corridors has an important role.However,from the matching degree of rail transit accessibility and car travel sharing rate,it is found that the utilization rate of cars along the track is not significantly lower than other regions,indicating that the replacement effect of rail transit on cars is lower.(2)Judging from the commuting methods of residents along the track,the proportion of rail transit commuting is 3.08%,slightly higher than the city’s average 1.3%,and the proportion of car commuting is 30%,which is consistent with the city’s average level.However,according to the statistics of the behavior space of the three types of residents along the line,it can be seen that the impact of rail transit on car travel varies significantly with the residents’ commuting space.Specifically,among the longdistance commuters whose commute OD points are located within the coverage of the rail station,the highest proportion of rail transit travel is 20%;only the long-distance commuters whose residence(point O)is located within the coverage of the rail station reach 11%;the long-distance commuters whose work place(point D)is located around the rail station have the lowest proportion of using rail transportation,only 5%.On the contrary,the choice of car travel mode is as high as 47%.(3)By dealing with the problem of spatial heterogeneity,the model results show that in the old city group sub-sample model,the variable of household income level is no longer significant for the choice of car mode,but it is significant in the suburban and remote suburban group samples.In general,after considering the spatial heterogeneity problem,the model variable saliency shows obvious sample differences,that is,in different sub-sample libraries,commuter car travel influence variables are different.Especially in terms of the impact of the built environment at both ends of the work and residence,the higher the built environment density at the end of the work site,the commuter car travel rate will be greatly reduced,and the rail transit has a stronger guiding effect,but the built environment at the end of the residence has a greater impact on commuters The impact of car travel is weak;secondly,residents in different locations have different socio-economic attributes,which leads to significant differences in the impact of rail transit on commuter car travel. |