Research On The Travel Potential Of Customiized Bus Based On Multi-source Data Fusion

As an important supplement to the conventional public transport system,customized buses have the advantages of rapid demand response,flexible route design and fast speed.The emergence of customized buses fills the gap in the coverage of conventional buses,improves the use efficiency of public transport and effectively attract private car owners to public transport.At present,the demand of customized bus mainly comes from customized bus platform combined with a certain amount of travel survey data.However,this demand prediction method fails to identify the potential users of customized buses effectively,resulting in the imbalance between supply and demand in many customized buses.Therefore,this paper proposes the potential analysis of customized public transport based on multi-source data to provide a theoretical basis for the development of customized buses.First of all,this paper establishes a bus passenger station identification method based on the multi-source data fusion.Structures of multi-source data and the correlation between the multi-source data are systematically described.In addition,the method of data preprocessing is introduced.The density clustering algorithm is used to fuse the static bus station data with the bus GPS data in spatial dimension to obtain the bus arrival schedule data.Then,the bus arrival schedule data and bus IC card data were fused in time dimension to obtain the bus passenger boarding station.An identification method of potential passengers of customized bus is established.By analyzing the temporal and spatial characteristics of bus passenger flow,the peak commuting time of passengers is obtained.On this basis,the identification of commuters is carried out.On strength of the analysis of bus commuters,a spatial clustering method is established to obtain commuters’ location.From the perspectives of bus lines and stations,this paper explains the correlation characteristics between customized buses and conventional buses.Based on the study of passenger transfer mechanism after the emergence of customized public transport,the potential demand of customized public transport is identified,which provides a theoretical basis for analyzing the potential demand of customized public transport.A complex network model of customized bus travel potential is constructed.According to the characteristics of the model,the degree,strength,connectivity and complementary graph efficiency of the network nodes are selected as evaluation indexes,and the topology structure of the customized bus travel potential network is analyzed.Taking the bus multi-source data of Hangzhou as an example,based on the identification results of bus passengers’ residence and the spatial and temporal distribution characteristics of passenger flow,a customized bus travel potential network is established.The rationality of the model is verified by scale-free test of the network.Finally,this paper proposes a method for setting up customized bus stations based on the analysis of network key sites.Based on the TOPSIS method of grey relation,the evaluation model of node importance points in the network of customized bus travel potential is established.The evaluation index system of the site is constructed and the weight calculation method of the evaluation index is expounded.The results of the case analysis demonstrate that the model can effectively identify the key sites of customized buses,which has important guiding significance for the site planning of customized buses.