The Studies Of Cellular Automata Models For Bidirectional Traffic Flow

Recently, with the development of social economy and the advancement ofpeople’s living standard, the surge in traffic flow has resulting in more serious issuesof traffic congestion and environmental pollution, causing the concern of manyscientists. On our country’s road transport, the worsening traffic congestion andfrequent traffic accidents aggravate the acuity of traffic problems. The reason is thatthe road driving conditions complex as a result of the numerous civil population anddiversiform traffic tools, on the other hand the backward management technology andlow use of transport facilities due to the shortage of transport infrastructure and lack ofscientific transportation planning.In this dissertation, after a brief introduction on the cellular automata theory,research mainly carried out around bidirectional traffic flow models. In our country,there are a number of low-grade highway transport systems. It is a complicated trafficstate with a variety of performance or size vehicles mixing driving on the rural roads.Thus, exploring the basic law of the bidirectional traffic flow will help to improvetraffic conditions and the overall planning of the transportation network, so hasimportant practical significance. This dissertation summarized as follows:First, we simulated a two-lane bidirectional traffic flow model with mixed vehicleproposed by Zheng Rong-Sen. Then, Based on the model’s lane changing rules, twonew lane changing rules of two-lane bidirectional traffic flow model related to samemaximum velocity were proposed by taking into account the driver’s turningcharacteristic, and the actual operation of vehicles. The fundamental diagrams oftraffic flow under two new lane changing rules were obtained. Result shows that,compared with single-lane model, models with two new lane changing rules couldenlarge the home lane traffic flow when the density on the opposite lane is small. Bychoosing the appropriate model parameters, simulation results are similar with theexperimental data.Second, based on the existing two-lane bidirectional traffic flow model, a novelcellular automata model for two-lane bidirectional traffic flow was proposed by takinginto account the new lane changing rule. It included the abolition of restrictions on thequeue of low-speed car and introduced a novel concept about the behind security gapof adjacent channel for passing car. The fundamental diagrams of the traffic flow were obtained with period boundary condition, and compared to that of the original model.Results show that both the emergence of slow car and the asymmetry of car density ontwo lanes have important influence on the mixing traffic flow.Third, based on the improved two-lane bidirectional traffic flow model, thefundamental diagrams of the traffic flow were obtained with open boundary condition.On full or partitive open boundary conditions, we analyze the characters of thetransition from free flow phase to maximum current phase, or from free flow phase tojamming phase, and inter-lane coupling effect on the traffic flow. Results show that,the characters and the critical point of the system have significant changed underdifferent boundary conditions. The traffic flow under different mixing ratio of vehicleshas similar basic properties, but different threshold.Fourth, with open boundary condition, based on the improved two-lanebidirectional traffic flow model, a novel cellular automata model for T-shapedintersection traffic flow was proposed with forward moving vehicles running by NSrules. The fundamental diagrams of the traffic flow in each channel were obtainedunder different injection probability. Results show that, intersection has affected alllanes of traffic flow. The impact of traffic flow for main roads on branch roads isstronger than branch roads on main roads. The critical points of injection probabilitywere different.Finally, we conclude and show the implications and future researches.