In recent years,dockless bike-sharing have been rapidly rising in China and become an important part of urban traffic system.However,the development of dockless bike-sharing is mainly driven by the Internet economy and has some blindness.The research on its travel characteristics,influence relationship,development orientation and other aspects needs to be further studied.Existing studies show that the development of dockless bike-sharing has a certain impact on the urban public transport system,and there is a relatively complex and close cooperative and competitive relationship.Therefore,the research on the interaction between dockless bike-sharing and urban public transport system is essential to clarify the functional positioning,development direction and development strategy of dockless bike-sharing.Based on the survey data of travel around XinJieKou district in NanJing,this paper conducts case study and effect evaluation,and carries out the research on the cooperative and competitive relationship between dockless bike-sharing and urban public transport system,which is interspersed in various chapters in a decentralized form.Firstly,combining the comparative analysis of the characteristics of dockless bike-sharing and public transport system,the competition and cooperation relationship between them are defined,and the cooperation mode is summarized.Combined with the travel survey data,the action form of cooperation and competition,the scope of influence and travel characteristics are elaborated,based on which the influence factors set of travel mode are preliminary selected.Secondly,based on the applicability analysis of Logit model,nested Logit model is selected to quantitatively analyze the influence of travel characteristics,personal attributes and other major factors on the choice of three usage modes of dockless bike-sharing.The results of model fitting show that the significance of the model is obvious,but the prediction hit rate of the model is relatively general.Then the algorithm model was constructed based on Adaboost-Nested Logit model.The concrete algorithmic flows of the Adaboost-Nested Logit model are designed.The structure and characteristic variables are determined to generate the improved weak classifier combination.In the case study,the established model is simulated by using the validation data set.Results show that the AdaBoost algorithm greatly improves the model hit ratio and has good application effect,providing an effective method for the construction of travel mode selection model.Finally,based on the analysis results of the cooperation and competition between dockless bike-sharing and public transport,the sustainable development model of urban public transport is studied.The functional positioning of dockless bike-sharing in public transport and corresponding development strategies are proposed to achieve the goal of sustainable and coordinated development of dockless bike-sharing and public transport system. |