Research On Multiple User Classes And Multi-Modal Stochastic User Equilibrium Model Based On Cumulative Prospect Theory

With the growing gap between supply and demand for urban transportation in China,traffic congestion has emerged as a major impediment to the growth of urban transportation in the country.To address the issue of urban traffic congestion,relevant government departments have developed appropriate transportation policies,enhanced transportation infrastructure building,and improved the degree of transportation demand management.A number of transportation modes have been deployed on the urban road traffic network to fulfill travel demand,offering travelers with more travel alternatives.In this thesis,we analyze the characteristics of multi-modal transportation networks,establish the topology of multi-modal transportation networks,investigate pedestrian decision making in road and transportation mode selection,and determine effective paths and design efficient path finding algorithms in multi-modal transportation systems based on interchange constraints.Based on the degree of travel network information and reliance on travel time reliability,travelers are divided into four categories:information-dependent travel conservative,information-dependent travel adventurous,experience-dependent travel conservative,and experience-dependent travel adventurous.The average travel costs of various sorts of travelers on each route section.This thesis proposes a cumulative prospect theory with generalized travel time cost as the endogenous reference point,constructs the value function and cumulative probability weight function of the path,calculates the path cumulative prospect value,proposes a path-aware Logit path selection model under the influence of effective path co-linearity,and takes the maximum cumulative prospect value as the basis for travel path selection.On this basis,the principle of multiuser class multi-mode stochastic user equilibrium is analyzed,and a multiuser class multimode stochastic user equilibrium model based on cumulative prospect theory is constructed.Using the construction method of Lagrange function to prove the equivalence and uniqueness of the model,the model is solved based on the Self-Regulated Averaging Schemes(SRA).In order to verify the difference of travel behavior of different travelers under the cumulative prospect theory of the model,a multimodal traffic network example is designed,and it is experimentally demonstrated that the degree of mastery of road network information and the degree of dependence on time reliability have a greater influence on the traveler’s traffic mode choice,and the effective path covariance has a greater influence on the traveler’s path choice.