Simulation Research On Complicated Coupling Dynamical Characteristics Of Traffic Flow

Traffic system is a complicated system consisted of many microscopic individual. There exist various accidental, stochastic and uncertain factors, such as individual behavior of driver, effect of unpredictable traffic incident and many nonlinear acting mechanisms in traffic system, which results in complicated relation among factors.With the rapid increasing of traffic flux, the interaction of vehicles is more and more obvious and represents strong coupling action. The coupling action mainly includes two sides which are the coupling between the front car and his following cars and between adjacent lanes. Relative change of one of factors is easy to induce change of others because of coupling action. It is effective means to study complicated traffic system by adopting traffic flow theory. And microscopic and macroscopic models are established to describe traffic flow. About modeling of traffic flow has many fruits. However, there are few studies on effect of traffic flow from coupling effect between cars. Therefore, it is not only helpful to understand deeply microscopic intrinsical characteristic and macroscopic propagating law under coupling effect but also a breakthrough point of traffic flow complicated mechanism by studying the dynamic characteristic that the coupling action impacts on traffic flow.In this dissertation, under the basic framework of traffic flow coupling problem being set up, along two lays expanding on microscopic car-following model and macroscopic dynamics model of traffic flow, a series of macroscopic and microscopic describing models of traffic flow are put forward according to two aspects of coupling effect between adjacent cars and between lanes. The effect of coupling effect between following car and preceding car to following car can be studied from microscopic car-follow model. Thus, the non-leaner dynamic feature for traffic flow, such as stability, non-leaner and hysteresis and so on, can be understood, which provides a novel angle of view to observe, analyze, understand and describe traffic flow phenomena. Secondly, the effect of coupling information disturbance and between lanes to traffic flow total wave can be understood from macroscopic model. Moreover the intrinsical law and real run mechanism of traffic flow are also discovered under coupling effect. The main contents of the dissertation are as follows:①New two car-following (TCF) model and multiple car-following (MCF) model of traffic flow are introduced for the coupling effect of preceding cars to following car caused by two cars and multiple cars coupling action in single lane, respectively.With the growth of modern traffic density, the effect from the running information change of non-adjacent preceding vehicles to the following car can not be ignored. The relationship of coupling interaction between vehicles is closer. Therefore, TCF model and MCF model are proposed according to the cases of two cars coupling and multiple cars coupling, respectively, from the effect of the information of multiple preceding vehicles to the following vehicle by introducing weight coefficient of coupling information. Analysis and simulation of our models indicate that the models can enhance the stability of traffic flow, make their delay time of car motion and kinematic wave speeds more approach to empirical value, overcome the phenomena of too high deceleration and avoid the disadvantage of negative velocity occurred at small sensitivity coefficientλin full velocity difference (FVD) model by adjusting the weight coefficient of coupling information. According to the linear stability analysis of the improved models, it is derived that it is the optimal state in MCF model by considering the coupling information effect of three preceding vehicles (m=3) to following car. That is to say, only the coupling information of three preceding leading vehicles (m=3) is enough to make the traffic become sufficiently stable under small disturbance.②Two new extended macroscopic dynamics models are proposed for the coupling effect caused by the effect of two cars coupling information disturbance and multiple cars coupling information disturbance to traffic flow in single lane.On the basis of microscopic car-following models, two extended macroscopic dynamics models for the disturbance of two cars coupling and multiple cars coupling have been deduced to study the effect of two cars coupling information disturbance and multiple cars coupling information disturbance to traffic flow by the relevant relation of microscopic and macroscopic scale parameter for the traffic flow according to the cases of two cars coupling disturbance and multiple cars coupling disturbance, respectively. The analysis and numerical simulation show that the improved model can reproduce the phenomena of jam and dissipation of real traffic flow. And the stable range of the improved model is enlarged obviously more than others by taking into account the disturbance of two cars coupling and multiple cars coupling.③Two-lane two delay time scales traffic flow coupling model and two-lane density viscous traffic flow coupling model are presented for the effect of the coupling effect and lane changing behavior between two adjacent lanes to traffic flow.On the basis of the two delay time scales continuum traffic model and Daganzo' model on multiple lanes, a two-lane two delay time scales traffic flow coupling model is extended to study the coupling and lane changing effects of the vehicles on two adjacent lanes by introducing coupling coefficient between lanes. The numerical simulation results show that the proposed model can correctly reproduce the coupling interaction and lane changing behavior between two adjacent lanes.However, it is non-anisotropic in multi-lane traffic because of lane changing. On the basis of the viscous continuum traffic flow model on single lane and Daganzo'model on multiple lanes, a two-lane density viscous coupling traffic flow model is proposed to further study the non-anisotropic characteristic of multi-lane traffic by introducing coupling coefficient between lanes. The analysis shows that this model is better capable of describing the non-anisotropic characteristic of real traffic caused by lane changing in multi-lane traffic.In this dissertation, a series of traffic flow macroscopic and microscopic models are established to cope with traffic flow coupling interaction effect relation through above researches. Furthermore, theoretical analyzing and corresponding simulation are implemented to study the traffic dynamics characteristic caused by coupling effect. Also, this dissertation validated the rationality of above models and obtained certain significant results which are more in accordance with the real traffic.