Travelers' Choice Behavior Of Bike-Sharing In Beijing Based On Hybrid Choice Model

With the acceleration of urbanization and the improvement of the level of mobility,the urban traffic problems in China are becoming more and more serious.In recent years,along with the era of "Internet + and shared economy",bike-sharing,as a new mode of travel and business form,has become the best choice for short-distance travel.In order to study the travel behavior of bike-sharing in depth,this paper takes the theory of planned behavior and technology acceptance model as the framework,establishes a hybrid choice model combining structural equation modeling and discrete choice model by introducing the latent variables of decision makers,and systematically and deeply analyses the personal attributes,travel attributes,environmental attributes and psychological factors on bike-sharing choice behavior based on the data of Beijing bike-sharing travel questionnaire.Firstly,Likert 7 scale,RP survey,SP survey and orthogonal design were used to design the questionnaire of bike-sharing travel.676 valid questionnaires were obtained by means of Wechat,QQ,E-mail,and interception survey in Beijing office buildings,shopping malls and other public places.Secondly,by constructing the structural equation modeling of bike-sharing use intention,the relationship between the latent variables such as perceived pleasure,perceived usefulness,subjective norms,behavioral attitudes and the behavioral intention to choose bike-sharing is analyzed.The results show that perceived pleasure,perceived usefulness,subjective norms and behavioral attitudes have significant effects on behavioral intention.Among them,perceived pleasure and subjective norms have indirect effects on behavioral intentions,perceived usefulness and behavioral attitudes have direct effects on behavioral intentions,and behavioral attitudes have the greatest impact on behavioral intentions.Thirdly,based on the questionnaire data,we select variables from four aspects:personal attributes,travel attributes,environmental attributes and traffic attributes to construct whether to choose bike-sharing Logistic model.The results show that teenagers,men,sunny days,holidays and other situations are more likely to choose bike-sharing travel,but with the increase of searching time and riding time,the possibility of choosing bike-sharing travel decreases.Finally,attitudinal latent variables are added to the traditional discrete choice model as explanatory variables to construct a hybrid choice model of bike-sharing choice behavior.The results show that the hybrid choice model improves the goodness of fit of the model and can explain clearly the impact of travelers' psychological factors on the actual choice behavior.A hybrid choice model was established for the samples of different travel periods,and it was found that the haze in the peak working days had a significant impact on the choice of bike-sharing,while the peak working days and holidays were not significant.When the holiday temperature is comfortable and the weather is clear,the possibility of choosing bike-sharing travel is greater,and people's behavior of choosing bike-sharing travel on holidays is vulnerable to the influence of others and the outside world.