Study On The Path Selection Of Urban Rail Transit Considering The Finite Rationality

Urban rail transit is an important part of urban transport.At present,it is showing a trend of large-scale and networked,which makes the optional routes in the passenger online network more diverse.The determination of the path selected by passengers is of great significance for the rational planning,operation organization and ticket clearing of urban rail transit.Therefore,how to scientifically predict and calculate the path selected by passengers has become a research institute for many scholars.However,among the existing two types of path determination methods,the first type of method is based on the logit model,aiming to describe the passenger path selection behavior by mathematical model,which has wide application range but lacks consideration of passengers' bounded rationality and heterogeneity of choice preferences.The second method is to mine and match the AFC card data,which neglects the behavior analysis,so the scope of application is limited.In view of the inadequacy of previous research,this paper takes the commuting passenger flow,the most important component of urban rail transit passenger flow,as the research object.It improves the first type of path estimation method and constructs a path selection model that is more in line with the passenger's actual choice behavior.It is expected to improve the accuracy under the premise of ensuring the scope of application of the model,and grasp the law of passenger path selection behavior at the same time.Based on the above-mentioned research objectives,the specific work contents and main conclusions of this paper are as follows:1.This paper summarizes the theoretical principles,theoretical models and theoretical applications of the prospect theory and the mixed logit model,which can reflect the bounded rationality and heterogeneity of choice preferences of passengers,and analyzes the connotation,behavioral mechanism and influencing factors of passengers' path selection behavior in urban rail transit,thus lay a theoretical foundation for the follow-up research.2.For urban rail transit commuter passenger flow,this paper designs a questionnaire for path selection behavior and conducts a path selection behavior survey.The questionnaire consists of two parts,one is the RP survey part that contains the passenger's personal economic characteristics and travel characteristics,and the other is the path selection SP survey part of the eighteenth groups of commuting situations designed by the sequential orthogonal design method.The implementation of the survey is based on a certain survey plan.It has requirements for the sample size,sample structure,and survey method.Finally,792 valid questionnaires were obtained,a total of 14256 sets of path selection observation data,which met the expected planning requirements,and provided a data basis for subsequent model calibration and selection behavior analysis.3.Considering the limited rationality of passengers and the heterogeneity of choice preferences,an improved model of urban rail transit commuter flow path selection is constructed.It also uses the survey data to perform parameter calibration of the model in this paper and previous research models separately.By comparing the analysis results,it shows that the model performs better in terms of attribute selection,goodness of fit and goodness of prediction,and can more accurately describe the path selection behavior of urban rail transit commuter flow,and its effectiveness is verified.4.According to the attribute parameters in the model calibration results,the path selection behavior of commuter passenger flow is preliminarily analyzed.The results show that the path feature attribute is the main influencing factor of passengers' choice behavior,and the preference heterogeneity of passenger does exist.At the same time,it found that the increase in age,income,and travel distance would reduce the passengers' perception of the path,and women were more likely than men.Secondly,through the marginal utility calculation of the path attribute parameters,it further analyzes the path selection behavior of the commuter passenger flow.It clarifies the specific impact on passenger route selection when the unit of each path attribute changes under different genders,ages,incomes,and travel distances.The results show that for commuter flow,the difference in gender,age and income has little effect on passenger route selection behavior.The overall ordering of the specific impact effects are transfer times,ride congestion,transfer travel time,ride time and transfer waiting time,but the sensitivity of different attribute groups to different path attribute changes is different.For the commuter passenger flow with different travel distances,the effect of different path attributes of passengers under short distance is different from the above sorting.Specifically,it is the number of transfers,the congestion of the ride,the travel time,the transfer travel time,and the waiting time for the transfer.This is due to the short travel time of short-distance passengers,and the sensitivity is also different.5.Through the sensitivity analysis of the subjective value parameters embodying the bounded rationality of passengers,it explores the law of the influence of different parameter values on the model effect.It separately obtains the values of the parameters under the goodness of fit target:,and the value of the parameter under the prediction goodness goal:.It also gives an analysis of the optimal parameters of the goodness of fit for passenger behavior and the use of predictive goodness optimal parameter values for the prediction of the use of passenger behavior.