| In order to alleviate the unbalanced distribution of inbound passenger flow during peak period and ensure the safety of passenger transport organization in urban rail transit(URT),station passenger flow control has become a basic operation management method.In practice,the implementation of passenger flow control measure at stations will increase the travel delay of passengers.Due to the high time value of the commuting passengers in the peak period,and the development of the integrated urban transportation network provides diversified choices of travel modes.Therefore,peak passenger flow control often brings response behaviors such as travel time adjustment and travel mode shift for commuting passengers,which cause changes in passenger flow demand,and affect the formulation of passenger flow control scheme.So it is particularly important to deeply analyze the influence relationship between control strategies and changes of commuting passengers’ behaviors.Based on these,this paper establishes a Bi-level programming model for peak passenger inflow control in urban rail network.When implementing passenger flow control scheme for rail transit stations,it also considers the demand changes of commuting passengers.Then,aiming to enhance the transport service level of URT during peak period and highlight the non-differentiated service concept of public transport,so as to obtain an efficiency-equality balance oriented passenger inflow control strategy during peak period in urban rail network.The main contents of this paper are as follows:Firstly,the survey on passenger travel behavior under the passenger flow control strategy is investigated.Combine the two survey methods of RP(Revealed Preference)and SP(Stated Preference)to collect the information on personal attributes,commuting statues,commuting intentions of URT commuting passengers.The RP survey is used to obtain the actual commuting travel time and travel path of commuting passengers.Combining with the scenario setting of passenger flow control strategy,the SP survey explores the decision factors of travel choice behaviors of URT commuting passengers,and reveals the time and space influence mechanisms on commuting passengers’ travel scheme choice under the implementation of passenger flow control strategy.Secondly,the discrete choice models of passenger travel under the passenger flow control strategy are constructed.The Baidu Map Web API(Web Application Program Interface)is used to extract the characteristic variables under different travel modes,such as travel time,travel cost.Based on SP survey data,the NL(Nested Logit)model of passenger travel time and travel mode joint choice under the passenger flow control strategy during the morning peak period is established to explore the travel behavior mechanisms of commuting passengers under the measure.Through the K-shortest path search algorithm,the minimum travel time and minimum transfer times are used as constraints to generate the effective travel path set of urban rail network.Based on RP survey data,the PSL(Path-size Logit)model of URT passenger travel path choice is established to clarify the influencing factors of passenger path choice,and provide necessary support for the formulation of passenger flow control scheme.Thirdly,a Bi-level programming model for peak passenger inflow control in urban rail network under changing demand is established.The upper-level programming model is a multi-objective passenger flow collaborative control model in urban rail network.On the basis of passenger arriving distribution and train capacity limitation,the upper-level programming model quantitatively expresses the dissemination process of passenger flow in the network,and the flow conservation constraints are established for the whole travel process contains platform waiting,train displacing,passenger transferring,etc.The goals of maximizing train boarding capacity utilization and minimizing passenger delay ratio difference at stations are constructed to propose a generation method of peak passenger inflow control strategy in urban rail network under the perspective of efficiency and equality.The PSL model can be used to calculate the basic indicators such as inbound passenger flow and transfer passenger flow in the direction,and then to solve the control strategy.The lower-level programming model is a passenger flow stochastic user equilibrium allocation model based on the travel choice NL model.By calculating the utility function and choice probability of various travel schemes,the dynamic changing state of passenger flow demand under the objective control strategy is obtained.Then,aiming at the constructed Bi-level programming model,a solving algorithm based on genetic algorithm and method of successive weight average algorithm is proposed.Finally,an empirical analysis of urban rail network in Guangzhou is carried out.By finding the optimal passenger inflow control scheme,then analyze the distribution of various indicators such as "train boarding capacity utilization ratio" and "passenger delay ratio at stations" in the example network before and after optimization.And the "full-load ratio" is taken as an auxiliary estimate indicator to reflect the effect of the proposed model and algorithm on alleviating congestion in the bottleneck interval.The results show that the passenger flow control scheme can effectively improve the utilization rate of train boarding capacity,and realize the equalization of passenger delay at stations,so as to ensure the efficiency and equality of urban rail system peak commuting service.Meanwhile,the maximum section full-load rate decreases from 126.6% to 106.9%,the Gini coefficient of full-load rate in the example network declines from 0.269 to 0.256,which shows that the local congestion phenomenon is significantly improved,and contributes to orderly and efficiently operation of urban rail network.Thus,this model can provide scientific instruction when implementing the work of passenger flow control in networking condition. |
PDF Full Text Request