Study On Passenger Flow Control Of Urban Rail Transit Based On Fluid Queuing Networ

Currently,congestion caused by dramatically increased travel demand is frequently occurring in urban rail systems.This reduces the level of service of the system and poses potential safety risks.The existing passenger flow control studies that give less thought to the interior of stations and ignore some characteristics of urban rail transit systems.In order to improve the service level and reduce the occurrence of congestion further,this study proposes an efficient two-level optimization method for collaborative passenger flow control at the network level and station level.The method incorporates four strategies: platform passenger flow control,train schedule optimization,service facility parameter control and outbound passenger flow control.(1)First,to address the lack of consideration of some characteristics of urban rail systems in existing studies,this paper integrates dynamic passenger demand,stochastic behavior of passengers,passenger interactions(state correlation),congestion propagation mechanisms and time-varying path selection probabilities with fluid queuing theory to lay the foundation for the subsequent modeling.(2)Subsequently,based on the analysis of the urban rail transit system,the stations are simplified into platforms in the network level,and the network level is constituted by platforms,boarding and alighting areas and trains;in the station level,the stations are decomposed into various types of service facilities,such as channels,escalators,stairs,security check systems and ticketing systems,and the station level is constituted by these five types of facilities.The corresponding queuing model is established for the above 7 types of facilities and structures with the help of the previously improved fluid queuing theory,and the queuing network model is built for the network level and station level respectively by combining the topology of the road network,so as to analyze the performance of the network.(3)In addition,based on the performance indicators obtained from the above model,we propose a two-layer optimization model that takes into account the service level,congestion and total passenger waiting time from both the operator’s and passenger’s perspectives.The network level of the model aims at minimizing the total passenger waiting time at platforms and concourses by implementing train schedule optimization and platform flow control strategies.Considering the difficulty of implementing platform passenger flow control,this paper further transfers the platform passenger flow control results obtained at the network level,i.e.,the number of passengers entering the platform at each moment,to the station level as a bridge between the upper and lower levels of the model.The station level passenger flow control optimization model uses station service facility adjustment and inbound passenger flow control strategies to minimize the passenger waiting time in the station facilities,while ensuring that the station output matches the platform passenger flow control results.(4)In order to effectively solve the proposed model,a solution framework based on the ADMM algorithm is proposed in this paper,firstly,the upper and lower level models are decomposed into two subproblems,and the continuous variable subproblem is solved by the improved nonlinear conjugate SPSA algorithm,and the integer variable subproblem is solved by the improved ALNS algorithm.Finally,the validity of the proposed method is verified by analysis based on a small-scale case and a real case.