Trajectories Estimation Of Passengers In Urban Rail Network Based On Smart Card Data

Uncovering the travel pattern and the flow distribution is important in the planning of timetable and responding of emergencies in urban rail network.Under the situation of "one ticket transfer",accurately and effectively finding the passenger travel trajectory in rail transit network,which is a black box,is the basis of passenger traffic management.Under this background of big data,it is of great significance for improving emergency response and planning of timetable to estimate passenger spatio-temporal location and get the distribution of passenger in rail transit network.Contactless smart card systems can record the passenger's travel information.Therefore,it is possible to estimate the trajectory of passenger by mining the travel information behind the AFC data.This paper first introduces the basic data of urban rail transit,and simply analyze the temporal and spatial distribution of passenger flow and introduces the characteristics of path' travel time based on AFC data.In order to estimate the trajectories of passengers,a passenger travel trajectory estimation model based on AFC data is proposed in this paper.Firstly,travel time model is established,on this basis,passenger travel time will be divided into the walking time between the charge gates and that platform,waiting time,in-vehicle time and transfer time.Different approaches are proposed to estimate the travel time elements by the typical passenger's tap-in and tap-out time in the URN.Then the travel time is used to match the passenger travel paths and find the path of passenger choose,and the passenger trajectory is estimated by the position estimation model.Finally,a single trajectory is mapped to the vehicle with the train timetable.In the case analysis,the time and space distribution in the URN,the number of passengers in the train and the number of passengers in the station are obtained,and the actual data is used to verify the model.The passenger travel trajectory estimation model based on AFC data is simple and easy to operate,but ignores the passengers in the actual trip process will be different,especially the waiting time and so on.On the basis of the former model,considering the preference of passengers,the travel trajectory estimation model based on passengers choose is proposed.Not only the passenger's waiting time is considered as a random variable,but also the proportion of choosing different paths is estimated in all possible paths.This model makes up for the deficiency of the previous model that does not take into account the preference of the passengers.In the example,the analysis and verification are carried out.