Research On Forecasting The Demand For Walking And Bicycle Traffic Based On Land Use And Built Environment

Pedestrian and bicycle transportation are gradually developed and widely concerned in China.Facing the new demand of refined design of pedestrian and bicycle transportation facilities,the regional demand prediction model represented by the traditional four-stage method can not fully and effectively respond to the refined planning and design of pedestrian and bicycle transportation and the decision-making needs of managers,while the micro level direct demand model(DDM)can make up for the shortage of the regional demand prediction model.The model uses regression analysis to quantify the correlation between the observed traffic volume and the characteristics of the surrounding environment to predict the pedestrian and bicycle traffic demand.In this paper,the direct demand model of stepwise linear regression is used to establish the micro level pedestrian and bicycle traffic demand prediction model based on land use and built environment,which has the advantages of low cost,fast response and easy interpretation of results.This paper focuses on the acquisition and processing of target data sets,measurement methods of indicators,model construction and other aspects.The main contents include:This paper summarizes four kinds of influencing factors of pedestrian and bicycle traffic demand(sociodemographic factors,land use and built environment,natural environment and time factors,concepts and attitudes);it focuses on the comparative analysis of various methods of pedestrian and bicycle traffic demand prediction,and confirms the feasibility of direct demand prediction model in micro level pedestrian and bicycle traffic demand prediction modeling Usability.Considering the availability of data resources,the importance of land use and built-up environment for shaping the behavior of pedestrian and bicycle traffic,this paper puts forward a micro level pedestrian and bicycle traffic demand prediction problem based on the perspective of land use and built-up environment,and analyzes the measurement index composition of multiple dimensions of land use and built-up environment.At the same time,considering the characteristics of pedestrian and bicycle traffic and the complexity of the dimensions of land use and built-up environment,the paper puts forward the problems that need to be solved in the model construction: obtaining,processing and index quantification methods of refined land use data suitable for pedestrian and bicycle traffic demand prediction;selecting appropriate variable space of land use and built-up environment Statistical range.Analyze the availability,timeliness and applicability of land use and built environment data sets,and use the data of network electronic maps as a refined source of land use data.The collection ideas and processing methods of network electronic map POI data,road vector data,and building vector data are proposed,and the multi-dimensional quantitative indicators of land use and built environment are measured by using Arcmap spatial statistical tools and the path planning interface of Gaode map open platform..Finally,taking the demand prediction modeling of pedestrian and bicycle crossing bridges as an example,Arcmap establishes a multiring buffer to select the appropriate statistical range of variables,and uses SPSS to perform manual stepwise linear regression analysis.Each variable is only detected at different spatial scales.It is allowed to choose once,and a theoretical model and a correction model containing explanatory variables of multiple spatial statistical scales are established.By testing the model,the results show that the overall accuracy of the model performs well.