Research On Recognition Of Abnormal State And Affected Transfer Passenger Between Hubs Based On Multi-source Data

With the gradual development of urban agglomeration integrated transportation network construction and the continuous expansion of intermodal transportation demand,the problem of transfer between multi-modal passenger transport hubs has become increasingly prominent.At the same time,all kinds of sudden abnormal conditions within the urban agglomeration will also bring impact to the multi-modal transportation network,so how to better distinguish abnormal events and identify the corresponding affected transfer passenger has become particularly important.This paper takes the travel between the integrated transport hubs within the Beijing-Tianjin-Hebei urban agglomeration as the research object,analyzes the identification method of abnormal passenger flow,and puts forward the identification method of the hub groups with strong correlation with abnormal condition and the affected transfer passenger.On the basis of analyzing and summarizing the types and identification methods of sudden abnormal states,this paper classifies the abnormal states occurring in the multi-mode transportation system of urban agglomeration,and puts forward the detection method of the abnormal aggregation of channel passenger flow based on Bayesian prediction,so as to obtain the safety threshold of passenger flow within a certain range.Then passenger travel information under normal and abnormal circumstances is obtained based on multi-source data such as mobile signaling data,stated preference survey data and ticket data.Among them,the mobile signaling data is obtained from the online database platform,and the Hive SQL statement is used to extract the passenger travel information within the research time and hub range.At the same time,combined with the ticket data and other information to investigate the passengers’ intention travel information under assumed abnormal circumstances,and analyze the attributes of the respondents and the choice of the scheme.On the basis of determining the validity and importance of association rules by using lift,cosine similarity and its standardized value,combining with passenger travel information under various abnormal conditions,association rules are used to analyze the relevance of hub groups under corresponding states.The influence range of the abnormal condition is determined according to the results of abnormal channel identification and the location of the assumed abnormal condition.Finally,the identification process of transfer passenger is given by using the time distribution sequence of transit chaining break and the information of passenger travel location,and the transfer time threshold of bus trip and intermodal transportation is obtained.At the same time,the different affected hub groups are determined to get the identification results of the affected transfer passenger,and the conceptual design of the relevant database is carried out.The study of this paper can determine the relevant information of affected transfer passenger between hubs under abnormal circumstances,so as to realize the active guarantee of transfer passenger between hubs,and provide support for improving the overall collection and evacuation capacity of urban agglomeration hubs and increasing the operating efficiency of the system.