Microscopic Traffic Simulation Considering Driving Characteristics

With the development of economy, traffic accidents, traffic congestion and othertransportation related problems become a significant problem in the world. In recentyears, traffic accidents happened more and more in China, and traffic jams isbecoming increasingly serious in some big cities such as Beijing and Shanghai.Therefore, how to take certain measures to solve these problems is a serious task. Theresearch on characteristics of drivers in China and microscopic traffic flow simulationcan provide the theoretical basis to accomplish the goal as above.After considering the risk-taking tendency and experiences of drivers, in thethesis, I classify all drivers into six categories, including risk-experienced diver,ordinary-experienced diver, conservative-experienced diver, risk-novice diver,ordinary-novice diver, and conservative-novice diver. Basing on literature review anddepth interview with Chinese drivers, I develop microscopic traffic performancemodels including the acceleration and lane changing models for each group. Theacceleration model defines two regimes of traffic flow: the car-following regime ifthe time headway is less than a threshold and the free-flow regime otherwise. Thelane changing decision process is modeled as a sequence of three steps: decision toconsider a lane change, choice of a target lane, and gap acceptance. The parameters ofthe acceleration models and lane changing models are measured though humorfactors experiments.Humor factors experiments are designed, and the data achieved by theexperiments are analyzed to determine all the parameters in the models. Then, amicroscopic traffic simulator is set up, and the acceleration and lane changing modelsare evaluated through their use in the simulator.Finally, using the simulator some researches on the factors which can exertinfluence on traffic flow was done, in order to offer help for traffic authorities whenthey make policies.