Research On Urban Rail Transit Passenger Guidance Method And System Under Congestion Condition

With the continues increasing scale of urban rail transit,the demand of passenger flow increases sharply,the contradiction between spply and demand in network intensifies seriously,decrease the quality of network service.How to alleviate the contradiction between supply and demand in time-space becomes one of the most important problems in the field of urban rail transit.Recent years,the urgent request for passenger flow guidance,especially for accurate guidance theory and practical instrucment for transit operator makes accurate guidance beomes a new hotspot.This paper focuses on passenger flow accurate guidance theory and practical method under congestion condition,researhes the scope of congestion in space-time,The distribution of affected passenger flow and guidance information under the passenger perception and its information optimization method.Based on these theories,this paper builds a system of urban rail transit passenger flow guidance based on network congestion identification.The main research contents of this paper are as follow:(1)Taking Guangzhou metro as an example,this paper analysis the congestion distribution in realworld,information release channel and level of released guidance information,and confirm the definition and neccisity of passenger flow guidance.On this basis,this paper decomposes the questions which can be used to achieve the accurate guidance and confirm the sub-question in the follow.Then,this paper analysis the critical factors which can affect the effect of passenger flow guidance from aspects of physics,demand and supply and confirm the concentration in the following research.(2)In terms of the methods to determine the scope of congestion in space-time:l)the key station identification model of urban rail transit network based on transfer network and spectral clustering is proposed,and gaussian kermel parameters and number of cluster are optimized with simulated annealing algorithm to optimize the effect of division;2)by mining operation data,constructing several indicators such as passenger flow contribution rate,information attenuation rate and configurable capacity coefficient,and combining with clustering analysis method,the congestion source station and secondary congestion station are analyzed and calibrated;3)by analyzing the historical operation data and determining the congestion regular of congested stations,this paper use the historical data to deduce the current congestion distribution at same period;Finally,on the basis of determining the scope of congestion in space-time,the affected passenger flow is demarcated,and empirical analysis is conducted.(3)In terms of release and optimization of guidance information,this paper firstly constructs a cluster analyzer based on passenger travel distribution,information attenuation rate and released level of guidance information to determine whether the affected passenger flow can perceive the guidance information.Furthermore,combined with MNL model,a passenger routing behavior model under guidance information is constructed.On this basis,a guidance information released and optimization model is proposed,and a solution algorithm based on genetic algorithm and simulation system is proposed.(4)Based on the theoretical results,an urban rail transit passenger flow guidance system based on road network congestion identification is built with ASP.NET MVC as the framework,and the design process of key modules is illustrated and show the interface of the system,achieving the transform from theorical to practical.