Study On Data-driven Bus And Bike-sharing Transfer

With the development of the city’s economic level,the per capita car ownership has increased sharply,resulting in increasingly serious urban congestion and air pollution problems.In order to alleviate traffic pressure and control air pollution from the root,the government has introduced the concept of green traffic and the planning policy of integrating public traffic with slow traffic.From the perspective of residents’ travel,conventional public transport is an important part of urban public transport,and bike-sharing is the most representative of slow traffic.However,when traveling in a short distance,bike-sharing takes the place of public transportation;When traveling for a long distance,changing to regular buses in bike-sharing can not only expand the service scope of public buses,but also play the service characteristics of "point-to-point" bicycles.The transfer form between bus and bicycle can effectively alleviate traffic congestion,which is in line with the concept of green travel,and has great research significance.However,due to the complete ride OD of rail transit in the existing research,the transfer research in bike-sharing mainly focuses on the transfer with rail transit,while most urban buses adopt the one-vote system,so it is relatively difficult to study the transfer between buses and bicycles.In order to better develop green transportation and promote the healthy and sustainable development of urban ground transportation system,this paper constructs NL model from the perspective of travelers to study the competitive relationship between public transportation and shared single workshop,and obtains the factors that affect the transfer between public transportation and bicycle.Based on the data of public transportation and bike-sharing,this paper studies the relationship between the number of transfer orders and the influencing factors,and finally puts forward the optimization measures according to the research conclusion.Firstly,through literature analysis,the factors that affect residents’ travel are summarized and analyzed.Through three aspects: personal attributes,travel attributes and travel environment,a questionnaire is designed,which lays the foundation for the research on the competition and cooperation relationship between public transportation and single workshop.Through the API interface of Gaode Map,the data of bus stops,routes,POI and traffic situation are collected,and the data are preprocessed by means of elimination and correction,so as to improve the accuracy of data fusion and matching,and provide data support for the identification of single car and bus transfer.Secondly,in order to ensure the independence of utility variables,the paper selects the influencing factors in the questionnaire through the tolerance characteristics of optimal scale theory.At the same time,in order to effectively alleviate the influence of IIA characteristics,NL model was used to quantitatively analyze the effect of selected utility variables on the choice of traffic travel mode,so as to characterize the competition and cooperation between transportation modes.The goodness of fit of the model was compared with that of MNL and NL models.The results show that the improved NL model has better goodness ratio and higher precision.Finally,by fusing the bus GPS data and the card data,the bus boarding station identification algorithm based on time information association is adopted.On this basis,an algorithm for recognizing the transfer behavior between single vehicle and public transportation is proposed.Through the correlation coefficient model,this paper studies the correlation between traffic situation,station transfer capacity,built environment and the number of transfer bicycle orders.The results show that the order number of transfer bicycles is positively correlated with the traffic situation and built environment,but negatively correlated with the bus transfer capacity.Finally,some reasonable suggestions are put forward according to the conclusion of the research on competition and correlation.In this paper,the optimal scaling theory is used to screen the influencing factors,improve the accuracy of the model,and make the study of residents’ travel choice more practical.The actual bus and bicycle data are used to study the transfer between bus and bicycle,which provides a basis for the planning of bus and bike-sharing transfer stations and routes.