Traffic Flow Simulation System Based On Cellular Automata Models

Traffic simulation is the technology to validate whether the improved methods are reasonable by reproducing, analyzing and evaluating the traffic condition, process and several management methods. The simulation result's precision depends on the traffic flow simulation models which are selected. Based on the cellular automata models in microscopic traffic flow simulation models, a traffic flow simulation system was designed for a city's road network.The simulation system's core module was divided into road segment and intersection's simulation module according to the vehicle's different running strategies on road network.The research which aimed at the module's different characters was done respectively.In the study of road segment's simulation module, in terms of dynamics of vehicle, road segment was divided into line road, road with courve and road with gradient to study how geometrical characteristic of different road segment influenced the vehicle's running. In the study of line road, the N-S model was improved; in the study of road with curves and road with gradient, new cellular automate models were put forward and tested in the simulation experiment. The simulation results showed that the new models were able to reflect the vehicle's running characters in special road segment, and proved the feasibility of the new models.In the study of intersection's module, a new approach to solve intersection's simulation was introduced, which absorbed the merits of one-dimension and two-dimension cellular automata models and combined the intersection's traffic flow organizing channel schemes. The difficult problem, conflicts in intersection, was studied with this new approach. The new approach can reflect the characters of vehicle running in the intersection and the experiment also shows it accorded with practical conditions and satisfied the requirement of intersection simulation basically.Finally, this paper designed the traffic flow micorscopic simulation system's functional modules which are based on cellular automata models. The modules' relationships and methods of realizing module function were also defined. Referred to the advanced technique of system integration, this paper gave suggestions on construction of system's module, which can be consulted for the advanced research.