Stochastic Network User Equilibrium With Route Choice Heterogeneity Under Adverse Weather Scenarios

Traffic network system is a typical uncertain,complex system,which will be inherently fluctuant with manifold uncertainties.Under these circumstances,applying traditional deterministic modeling approaches to describe the networks is generally insufficient to reveal the nature of realistic traffic activities.These unavoidable uncertainties of traffic networks logically derived from supply side(roadway capacity fluctuation)and demand side(travel demand variation).Adverse weather condition is one of the inducements that lead to supply uncertainty.On the other side,travelers' individually behavioral uncertainty,or to be specific,route choice heterogeneity,is considered as a non-ignorable reason resulting in variation of travel demand pattern,which is logically classified into demand uncertainty in this study.This dissertation proposed a novel stochastic network(deterministic)UE model that considers not only the different effects of the AWSs s on the degradable roadway capacity and various route choice criteria of travelers,but inserted by the effect of heterogeneous distribution of TBI on EOP.A GBPR function is adopted in this dissertation to model the negative effects of AWSs on free flow speed and roadway capacity.Based on the impact of AWSs,the considerations of route choice heterogeneity of travelers in this dissertation are twofold.The population is divided into TPI and TBI by their various network information levels(NILs)and,the TBI are further classified into THeBI by their different scenario knowledge levels(SKLs).Based on the discrepancy between TPI and TBI,ARTT and PSRTT are selected as the route choice criteria for TPI who are armed with ATIS and for TBI under uncertainty.Moreover,this dissertation develops the adverse scenario-based stochastic network-deterministic user equilibrium(AWS-SN-DUE)traffic assignment models.The contents of this dissertation include the following aspects on:(1)impact analysis of the adverse weather on roadway transportation;(2)the definition of route choice heterogenerity under adverse weather conditions in this dissertation;(3)two single-user-class SN-UE problems under adverse weather scenarios based on the route choice homogeneity;(4)two multiple-user-class SN-UE problems under adverse weather scenarios based on the route choice heterogeneity;(5)numerical study based on practical case.The findings of this dissertation are listed as follows:(1)In S-TPI-UE model,the route travel time raises with the increase of route flow and rainfall intensity.For S-THoBI-UE model,with the increase of rainfall intensity,THoBI are more intended to choose those routes that are more steady with adverse weather scenarios.More accurate network information will not necessarily enhance the performance of the whole network.(2)As the AWS deteriorates,travelers tend to prefer weather-proof routes to those weather-sensitive routes,and higher percentage of travelers in developed areas select weather-proof routes than those in developing areas do.THoBI are more evenly distributed on the network when they have to make decisions by their own traveling experience without the help of ATIS.(3)The advantage of SDA is its theoretical similarity to Frank-Wolfe algorithm that is easy to be implemented.However,the numerical examples also testify that the shortage of SDA is that it converges fast at the beginnings of the algorithms but acts dilatorily afterwards.(4)When the network is in its most adverse scenario,THoBI may underestimate the ARTT of TPI,while the scenario is perfect,THoBI may overestimate the ARTT of TPI.(5)When the network scenario is perfect and so is the network capacity,ATIS plays positive role in guiding TPI and those THoBI who consider the EOP is accurate to high-capacity routes to lower the route travel time.However,when the network scenario is more and more terrible and so is the network capacity,ATIS tends to guide TPI and those THoBI who consider the EOP is accurate to assemble on those weather-proof routes to worsen the overall performance of the system.