Study On Demand Prediction Analysis Technology Of Rapid Urban Transit System Based On Combinatorial Optimization

Traffic congestion is becoming more and more serious in metropolitans.In order to solve this problem,the urban passenger transit system should be firstly studied to form the structure of urban passenger transport with rapid transit as framework;normal bus as the main part and taxi,etc to supplement.Rapid bus is punctual,comfortable and high-capacity on behalf of the development direction of future traffic and also one effective way of urban traffic problems.The demand forecasting technology is given to support the rapid bus planning.Then,on the basis of deep analysis of relevant literature at home and abroad,rapid passenger transport system is described,the relevant factors which impact the demand forecast,regional sub-impedance during the distribution phase,the share rate of rapid bus,the path selection in super-network and traffic assignment are analyses focally.One joint prediction model is put forward based on the four-stage prediction model and the equivalence and uniqueness are proved.The main contents of this dissertation include as follows:1.On the basis of urban transit system frame analysis,the concept of rapid passenger transit system is put forward.This paper is limited in the scope of rapid bus as the core.The relevant factors which impact the demand forecast of rapid passenger transport system are analyzed deeply and the principles and overall plan are put forward.2.During the four-step prediction method,generalized path costs are computed respectively by the basis of road network and rapid bus network and the result can be used for the calculation of regional impedance matrix for the distribution forecast of passenger flow.3.The three-stage classification approach is put forward according to the different features of modes selecting to calculate the distribution volume of public transport and the transit share rate of rapid bus is calculated combined with the special natures.4.The route choice behavior of passenger is analysis according to the theory of total lines,strategies,and super-path and the traffic distribution is performed in the super-network considering both cases of Capacity constraints and non-restricted.5.A joint model of passenger flow distribution,mode split and passenger flow assignment is established based on the four-stage method which according to repeated adjustments in the inter-phase to achieve the balance between section flow and ultimate way through a very small first-order optimality conditions of Lagrange function to prove the equivalence and uniqueness of the model.Finally,a numerical example of Beijing rapid transit planning through Fushi road is taken to validate the rationality and feasibility of the theory and method with the external development software TransCAD.