Dynamic Traffic Characteristics And Traffic Assignment Method On Urban Road Under Rainstorm Weather

The urban road transportation system,as the skeleton of a city,is the carrier of passenger and freight transportation,social and economic activities,and citizen lives.At the same time,it also bears the threat of natural disasters.Frequent rainstorms severely affect the operation of urban road traffic and transportation,reduce the traffic supply capacity,and aggravate traffic congestions.Existing studies usually focus on the overall effect of rainstorms and ignore the dynamic rainfall conditions during a rainfall process,so it is difficult to describe the dynamic relationships between traffic characteristics and rainstorms accurately during the rainfall process.Moreover,the dynamic temporal and spatial distributions of traffic flow on the urban road network under rainstorms need further analysis.Therefore,it is necessary to in-depth analyze the dynamic degradation of urban road traffic characteristics under rainstorm weather,grasp its dynamic correlations with rainfall,road and vehicle factors,and further study the traffic flow assignment model and algorithm in dynamic rainfall conditions,to provide theoretical support for traffic management and intervention countermeasures in response to rainstorms.Based on the dynamic traffic characteristics of the urban road under rainstorm weather,this dissertation investigated the dynamic degrading evolution and the Dynamic Traffic Assignment(DTA)method on urban road network under rainstorm weather.First,the comparison analysis of traffic flow characteristics under rainstorm and normal weather was conducted based on the historical traffic data and rainfall data in Shenzhen.Then,the rainfall impacts on the fundamental diagram of traffic flow and the traffic congestion were discussed from the perspective of the rainfall process,to clarify the influence mechanism of rainstorm weather on urban road traffic.Furthermore,extracting the dynamic characteristics of rainfall intensity and sliding rainfall from the rainfall process,this dissertation analyzed the influencing factors of free-flow speed degradation and established a dynamic degradation model of free-flow speed under rainstorms.The evolution direction of urban traffic characteristics under rainstorm weather was revealed.Finally,the travel cost function under rainstorm weather was built by combining the physical queuing model with the dynamic degradation model of free-flow speed.Moreover,the DTA model for urban road network under rainstorms was established based on the principle of Stochastic Dynamic User Equilibrium(SDUE).The research content mainly includes four aspects:(1)Taking urban road traffic as the research object and selecting rainstorm as the typical severe weather,this dissertation first gave the classification criteria of the rainstorm weather.Then the rainstorm analysis object and the normal weather control group were selected according to a certain principle,based on the historical traffic data and rainfall data.After the processing and converting of the original data,the traffic flow characteristics distributions under the normal weather and rainstorm weather were compared and analyzed.(2)Taking the rainfall process as a research point,this dissertation proposed the importance of analyzing the influence of the rainfall process on traffic flow under rainstorm weather.That is based on the influence mechanism of the rainstorm weather on urban road traffic.The phase division method of the rainfall process is given,and the rainfall process was divided into pre-peak and post-peak stages according to the research purpose and data conditions.Based on the traffic data measured on the major arterial roads and the rainstorm data,the Underwood model was used to generate fundamental diagram models at rainstorm pre-peak and post-peak stages,respectively.The influence difference of rainfall process on traffic flow at different stages was revealed.Then,from the perspective of the traffic congestion formation mechanism,the congestion elements for analysis were determined.The traffic congestions at rainstorm pre-peak and post-peak stages were compared and analyzed.The generation of occasional congestion and the aggravation of frequent congestion under rainstorm weather were explored.This study makes up for the shortage of lacking the influence analysis of rainfall process on traffic in existing studies.(3)From two perspectives of the theoretical model and measured data,this dissertation explored the methods to obtain the free-flow speed of expressway under normal weather.Concentrating on the vehicles in the free-flow state,the research premise for the free-flow speed degradation in rainstorm weather was provided.The free-flow speed degradation ratio was confirmed by comparing and analyzing the free-flow speed under normal weather and rainstorm weather.At the same time,and the influence degree of rainfall intensity,sliding rainfall,lane location,vehicle type,and water-accumulating section type on the degradation of free-flow speed was explored.Based on the dynamic characteristics of rainfall intensity and sliding rainfall and considering their interaction with categorical variables,the method of introducing categorical variables into the model was proposed and simplified.A dynamic degradation model of free-flow speed under rainstorm weather was established.Based on the measured data,the accuracy of the model was tested through the verification set to determine the optimal model and related parameters.The research results solved the problem of less considering the issue of rainfall dynamics in current free-flow speed degradation research.The model results and parameters can be applied to the traffic flow analysis of expressways under rainstorms and the real-time prediction of the dynamic degradation ratio of free-flow speed.That will have positive effects on the development of road degradation research under rainstorm weather.(4)Based on the rule of dynamic degradation of free-flow speed in rainstorm weather,the dynamic rainstorm factor was considered in the travel cost function of road sections.Combined with factors such as road traffic load,vehicle type,and road section type,the travel cost function of road sections under rainstorm weather was established based on the physical queuing model.That is to reveal the travel cost composition for different types of vehicles and realize the dynamic estimation of the travel cost for degraded road sections under rainstorms.Then,taking the path-sensing instantaneous travel cost as the basis for travel decision-making and choosing SDUE as the DTA principle,the departure time and route choice models under the influence of rainstorms were proposed to clarify the propagation law of traffic flow.The DTA model and algorithm under rainstorm weather were constructed,which solved the problem of lacking the analysis of dynamic rainfall conditions in existing DTA studies on degraded road networks.Through a case study,the temporal and spatial distributions of path flow under different scenarios were discussed.The results show that our model can effectively analyze the dynamic traffic flow assignment under rainstorm weather and explain the abrupt and gradual changes in path flow at different stages of the rainfall process.The research results can provide a strong theoretical basis for traffic emergency management under rainstorm weather.The research results of this dissertation enrich the theories and methods in the field of traffic flow assignment under severe weather and provide guidance for decision-making departments,to master the temporal and spatial distributions of road network traffic under rainstorm weather and optimize traffic emergency management and evacuation strategies.