A Dynamic Stochastic Disequilibrium Passenger Flow Assignment Method And Its Application On A Local Disruption Of Urban Railway Network

With the rapid development of China’s Urban Mass Transit, the network structure is becoming more and more complex, daily passenger flow keeps rising, and emergencies happen frequently, which bring the operation department a huge challenge on the safety of passenger flow organization and emergency disposal. So it has become an urgent need to find a method which can describe the passenger flow distribution dynamically under different conditions.On the basis of a further research of the traffic flow assignment related theories, we proposed a dynamic stochastic disequilibrium passenger flow assignment method, which could not only afford the need to make an assignment of daily passenger flow under normal conditions, but also calculate the passenger flow distribution when the network has a local disruption.Firstly, this paper made an analysis of the disequilibrium of Urban Railway Network, summarized the causes; then discussed the general steps and related theories of traffic flow assignment, built up an Urban Railway Network topology model and a travel impedance model on the basis of the travel path dual graph under comprehensive consideration of the cases including interchange, crowd and congestion. Afterwards, we defined the valid travel path of urban mass transit more comprehensive and detailed, proposed a breadth first valid travel path search algorithm, built up the travel utility function and improved the multipath passenger flow distribution logit model as well as a passenger flow index system consists of3layers including station, line and network, and proposed the related calculation method under time constraint.On the basis of which, we researched the application of the above model and algorithm under a local section disruption condition, summarized train operation plan under fault conditions, made an classification of passenger according to the characteristics of them, proposed the method to calculate the aggregation number of the station in case of insufficient transport capacity, introduced the passenger flow distribution method under local section disruption condition.At last, with programming language, we simulated the dynamic propagation process of the working day morning peak passenger flow on Beijing Urban Rail Transit network with real data. Proved the efficiency and validity of the proposed model and algorithm, which provided a theoretical basis for the urban mass transit emergency management.