Research On Road Traffic Flow Simulation Model Construction And Its Appliaction

Traffic simulation originates from the researches of traffic flow theory in the1950s, is an interdisciplinary combining system simulation technologies andtransportation science. After half a century’s development, it has been one of the mostimportant branches of transportation management research field. The5factors topromote the development of traffic simulation modelling research include: traffic flowtheory, computer software technology, computer hardware technology, informationtechnology infrastructures and transportation management demand. Currently, allthese factors are in rapid development. Especially in recent years, the rapiddevelopment of information technology makes more complex traffic simulationmodeling possible, traffic management, planning and decision-making also has putforward higher requirements on the simulation granularity, efficiency and scalabilityof traffic simulation. Therefore, taking full advantage of the progress of systemsimulation technology and computer processing technology, building anindependentand controllable road traffic simulation model to support the analysis of variousproblems stem from transprotation development has important practical andtheoretical significance.In this dissertation, from three levels including links, static road network, anddynamic road network, road traffic simulation modelling is strudyed on the basis ofMonte Carlo simulation and discrete event simulation method. The main contentsinclude: research on road traffic flow simulation model construction and systemimplementation considering the difference of different vehicles’s behaviors; acomplex-network-based static road network simulation modeling and systemimplementation; dynamic road network simulation modeling and systemimplementation based on O-D estimation method. Three levels of simulation modelstogether to form unified road traffic flow simulation model. By taking road trafficpollution emissions and road congestion management strategy analysis as case study,this paper further to study the application of the simulation models.In the study of link traffic flow simulation modelling, this thesis proposed amicroscopic traffic flow simulation system considering the difference of individualvehicles’s behaviors based on Monte Carlo simulation method. The characteristics oftraffic flow on roadways are obtained by counting the characteristics of vehiclestraveling on these roads. The simulation characteristics of vehicles are presented by aset of variables. These variables change in the way of Markov process in simulationprocess. The simulation system is advanced by discrete-time mechanism. Thesimulation system realizes the differentiated simulation of individual vehicles’behaviors by introducing random variables; and the simulation model can be easily extended to enable it to be applied in the analysis of specific traffic problems.In the construction of static road network traffic flow simulation model, thisthesis introduces the concept of link resistance into the complex network theory toconstruct a static road network traffic simulation model capable of taking into accountboth traffic flow on links and the topology structure of road network. A lot oftransportation network measurements results show that road networks in China havesignificant complex network structural features. The present researches focus on thetopology features of road network and lack the attention to the relationship betweenlink traffic flow and road network topology. This thesis introduces the concept of linkresistance into road network topology simulation modellying to realize the simulationof the influence of traffic flow on road network topology. Link resistance is the timedelay through a link, depending on link design features and the traffic flow on link.Links correspond to edges in complex network, and accordingly link resistance can beseen as the weights of edges. Changes in traffic flow will influence link resistance,thereby influence the weights of edges in road network and making road networktopology changes. After the construction of the road network simulation model, thisdissertation also presents a road network congestion simulation evluation methodbased on the time of virtual test vehicles travesing the whole road network, andrealizes this simulation system.For the construction of dynamic road network traffic flow simulation model,based on the above static road network traffic flow simulation model, this thesisfurther developed a dynamic simulation model of road network traffic flow which isable to consider travelers’ travel demand. By the introduction of the gravity-basedtravel demand estimation model into simulation system, combined with theshortest-resistance-path selection method, the simulation model realizes the mappingfrom travel demand to traffic flow in road network. The dissertation also realizes thesimulation system on the basis of the discrete-event simulation method; as opposed tothe existing road network traffic flow simulation systems, this road network trafficflow simulation system can simulate the complete dynamic evolution process fromsocial and economic factors to road network traffic flow.Finally, we choose traffic emissions and traffic congestion management, the twoaspects of most concerns in transportation management field, to carry out theapplication studys of the simulation models we constructed. By the introduction ofvehicular emission factor into vehicle model, a link traffic emissions simulationmodel is presented. Road network traffic emissions can be obtained by adding theresults of all link emissions together.In addition, whether tidal lanes or trafficcongestion fees, the nature of these traffic management measures can be seen astraffic management measures based on road network topology control. The smoothimplementation of the traffic management measures based on network topology control requires a precise understanding of the complex relationships between theroad network’s topology structure and its traffic carrying capacity. This dissertationuses the road network traffic simulation model we constructed before tosystematically study the complex relationships between the road network’s topologystructure and its traffic carrying capacity in terms of accessibility and unimpededcapability of transportation network topology structure. Reasearch results are helpfulfor the design, decision-making and analysis of traffic congestion managementmeasures based on road network topology control.