Simulation of traffic flow on the basis of Road Network-Based Cellular Automata

In order to promote the use of simulation in the traffic navigation and management, the behaviors and existing simulation models of the moving objects (such as micro-cellular automaton model, Poisson distribution of microscopic traffic simulation model and section start.) in traffic flow have been researched, integrated and improved in this paper, and provides the Road Network-Based Cellular Automata Model (RN-CA) in eventually. This model not only simulates the behaviors of the moving objects from the angle of vehicle and road, but also adds the departure, following, overtaking and correction model in different types of car and the major factors that affect vehicle operations, making simulation more accurate and reliable. This paper using the Dijkastra algorithm, and considering the length of the road, the number of lanes, traffic, average speed and other factors to select the optimal path. Besides, the optimal path can change according to the road conditions and time. For driving time estimation, this paper combines the simulation and prediction, considering the traffic of current and historical. The prediction process uses the changing polynomial fitting and the interpolation fitting. In order to adapt to Multi-core computing, the system provides a simulation mode that can adjust threads by users, and the efficiency of the system is improved.