Study On The Model Of Traffic Flow Based On Cellular Automata Theory

Cellular automata model is an important mathematical model of traffic flow theory, itstime and space are discrete which making use of simple rules to simulate the complex trafficphenomena. As the same time, it computes efficiently and implements easily. Therefore, it hasalways been a hot topic of traffic flow and has a broad application prospects.On the basis of the existing cellular automata theory, this paper puts forward someimproved models, then with the models to simulate and analysis the traffic flow, and with it tostudy the complex characteristics of traffic flow. In particular, it mainly includes the followingcontents:(1) It introduces the basic concepts of traffic flow theory and cellular automata, studiesthe current some mainstream cellular automata traffic flow models and classifies these modelsfrom a single-lane, two-lane and the transition rule of view. Then combined with the measuredtraffic phenomenon, it analyzes the advantages and disadvantages of each model from thebasic diagram of traffic flow and the space-time diagram of traffic flow.(2) On the basis of single-lane NaSch model, this paper establishes an improved model.This new model considers the real-time effects of the front car in the acceleration and brakingstep, at the same time, its randomization deceleration probability is a function of vehicle’sspeed and density. Comparing the simulation results with actual measured data, the resultshows that this improved model under periodic boundary conditions can simulate the moremoderate steady driving behavior than NaSch model. Then this paper extends this improvedsignal-lane model to two-lane, the simulation shows that it can better describe the phasetransition process that road traffic flow from free flow to the crowded flow.(3) With respect to the mixed traffic in realistic road traffic, we respectively set up mixedmulti-speed single-lane and two-lane cellular automata traffic flow model, the simulationshows that the mixed traffic flow and speed of traffic flow is basically determined by the slowcar’s flow and speed. Then considering different performance vehicle should have moderatedprobability, this paper establishes an improved model that is based on vehicle performance,using the improved model to simulate them, the result shows that this improved model ofmixed vehicle traffic can solve the problem that slow vehicle effect is overestimated.Comparing with the previous model, the simulation results show that the model that isbuilt by this paper is close to the actual traffic flow, providing certain scientific basis for thestudy of the actual road traffic conditions.