Road Net Construction And Vehicle Behavior Modelling In Microscopic Traffic Simulation

With more realistic simulation ability of real vehicle behavior, capable to provide differentlevels or kinds of traffic information and good at displaying the effect of traffic measures bydynamic computer graphics, Microscopic Traffic Simulation becomes a very useful andeffective tool in solving traffic problems. However, Microscopic Traffic Simulation can onlygive the practical reference result with accurate and completed road net data plus effectivevehicle behavior model. So the focus of study in Microscopic Traffic Simulation lies in roadnet construction and vehicle behavioral model.In order to improve the description ability and construction efficiency of road net, thisthesis proposes a3-level road net model and its construction method based on Google Map.This road net model can be organized by different zoom level from road, link and lane. What’smore, it not only consists the geometric data which can be used in3D environment to bulidroad, but also consists the logic information of road net. Using the google map method givesmore autonomous in construction. Construction method based on SHP is also given which savestime in constructing a large scale of traffic road net.In addition, in order to simulate vehicle behavior more realistically, the study starts fromthe reason and feature of vehicle behavior. Then dividing vehicle behavior into two types andmodelling respectively, the two models are Following-leader model and Lane-changing model.Following-leader model can describe vehicle behavior more realistic by extending the IDMmodel, Lane-changing model supports continues lane-changing behavior and trace in3Dsimulation environment. By siting some personalized parameters, vehicle behavior model canincrease the diversity of simulation results.Lastly, this thesis gives the software design and implementation of real traffic simulationsoftware with road net construction method and vehicle behavior model based on the methodand definition proposed in thesis. Furthermore, these models’ practical effect are verified bytesting this traffic simulation software.