Analyzing Of Typical Problems Of Urban Road Traffic Flow Based On Microscopic Models

Traffic congestion is one of the most serious problems in most cities all over the world. It restricts the developing of a city and social economy to healthy direction. Therefore, one important problem should be charged by traffic flow theory is to deeply understand the natural properties of traffic flow and the cause of traffic congestion. Appropriate traffic flow models can thus be proposed, and traffic management and control strategies can also be designed according to the theoretical results. So far, traffic flow theory has developed vehicle traffic, although there are still lots of traffic problems should be solved On the one hand, most works of traffic flow theory focus on the single vehicle traffic. In most cities of our country, the major characteristics of traffic flow are mixed traffic and low velocity. It is much more difficult for the modeling of this inhomogeneous flow. On the other hand, with the rapid development of intelligent transport system (ITS) in recent years, the modeling of vehicles with ITS, as well as the effect of vehicle with ITS on mixed traffic flow are paid attention to.Based on problems mentioned above, two subjects of traffic flow are investigated in this dissertation. Firstly, in view of the basic characteristics of urban mixed traffic of our country, the interaction mechanism among mororized vehicle (m-vehicle), nonmotorized vehicle (nm-vehicle) and pedestrian is investigated. The microscopic modeling of road, interaction and crosswalk (where the mixed traffic is serious) is performed respectively, and numerical simulations are carried out to investigate the characteristics of mixed traffic flow. Secondly, considering the rapid development and application of intelligent transportation system (ITS), an extended car-following model is presented by incorporating the ITS information. Further more, the effects of vehicles with ITS on the traffic flow of a one-lane road as well as an on-ramp system are also investigated. The contents of this dissertation are as follows:(1) A coupling cellular automaton model is proposed to depict mixed traffic flow of a road without isolations for segregating nm-vehicles from m-vehicles. The model is combined by the NaSch model and the multi-value cellular automaton model. It takes into account both the lane based motion of m-vehicle and non-lane based characteristics of nm-vehicles. In addition, the rules for avoiding collision and gridlock are also presented to depict the interference between m-vehicles and nm-vehicles. Simulation results show that both first-order and second-order phase transitions can be obtained for m-vehicle flow, corresponding to small and large densities of nm-vehicle, respectively. The flux curves of m-vehicle become lower as the density of nm-vehicle increases. The maximum flux and the corresponding density of nm-vehicle flow decrease first and keep constant later with the increase of m-vehicle density.(2) Based on the two-dimensional optimal velocity model, a new two-dimensional car-following model is presented by taking into account the effect of velocity difference to investigate the motion of m-vehicle and nm-vehicle near an typical unsignalized intersection. The simulation results show that the straight going m-vehicle flow just next to nm-lane is disturbed more seriously than others. At a small entrance gap of m-vehicle, the flux of m-vehicle (nm-vehicle) decreases monotonously as the left-turning probability of nm-vehicle decreases, while there is a critical point on each flux curve at a large entrance gap of m-vehicle, which divides m-vehicle (nm-vehicle) flow into free flow part and congested flow part. In addition, a well-known phenomenon in reality. is observed that groups of m-vehicles and nm-vehicles pass through the intersection alternately.(3) A coupling cellular automaton model is presented to depict a road with a signalized crosswalk, and simulations are carried out. In simulations, pedestrians are classified into three types:careful pedestrians who obey traffic rules, riskers who do not obey traffic rules, and normal pedestrians who obey traffic rules when their waiting time is less than a critical value and do not obey traffic rules otherwise. The case for pedestrians consisting of careful and normal pedestrians is investigated by simulations. It can be found there is one critical point on each flux curve of vehicle flow which divides the vehicle flow into free flow part and congested flow part. While there are two critical points on each flux curve of pedestrian flow, which divide the pedestrian flow into three types, that is, free flow, congested flow and medial state (both free flow and congested flow can be found in a signal period). The effect of proportion of normal pedestrian and the length of signal period are also investigated by simulations. Compared with normal pedestrians, riskers may cross the road at any time in a signal period, and so they contribute much more interference to vehicle flow. Finally, the frequency variation for riskers crossing the crosswalk during a signal period is investigated. It can be found when pedestrian flow consists of all the three types of pedestrians, there are two local maximal frequencies. This is a typical self-organization phenomenon observed in real traffic.(4) Based on the Optimal Velocity (OV) model and its extended models, the multiple headway and velocity difference (MHVD) model is proposed by taking into account the effect of ITS information, which includes both the headway and velocity difference of multiple preceding cars. The stability condition of the model is obtained by using the linear stability theory. Through nonlinear analysis, the mKdV equation is constructed and solved. According to the neutral stable line and the coexisting line, traffic flow is classified into three types, i.e., stable, metastable, and unstable. The simulation result shows that the traffic jams can be suppressed more effectively by taking into account both ITS information of headway and velocity difference than taking into account only one of them, which is in good agreement with the analytical ones.(5) Effect of vehicles with ITS on mixed traffic flow is investigated by simulations. The OV model is used to depict the motion of normal vehicles, and the MHVD model is used to depict the motion of vehicle with ITS. Firstly, the fundamental diagram is investigated for various of proportion of vehicle with ITS under periodical boundary condition. It can be found vehicles with ITS make the traffic flow more stable at medial density region, while they have little influence on traffic flow at small or large density region. In addition, the improvement induced by vehicles with ITS decreases with the increasing proportion of vehicles with ITS. Secondly, under open boundary condition, the effect of vehicles with ITS on traffic flow near an on-ramp system is investigated by introducing rules for merging vehicles from the on-ramp to the main road. Simulation results show that the free flow region of main road is enlarged by vehicles with ITS, and the congestions are suppressed. However, congestions on the on-ramp lane become more serious. From simulations, it can also be found that the maximum saturated flux of the on-ramp system is promoted by vehicles with ITS.