Research On Intelligent Traffic Simulation Based On Agent And Cellular Automata

Traffic problem is a global hot issue that has seriously influenced our society. With the rapid progress of computer science, traffic simulation has become an important means to research and solve traffic problem. Many traffic simulation models had been proposed in the past forties years, and this thesis first puts forward a new method of traffic microsimulation which is based on the combination of Agent and Cellular Automata theories.Firstly, the thesis analyzes the characteristics of two methods of traffic simulation that based on Multi-Agent System (MAS) and Cellular Automata (CA) separately, and the feasibility of combining these two methods. Then, based on some skillful extensions and modifications of the two methods, this thesis puts forward a traffic microsimulation model based on the combination of MAS and CA. The architecture of the model is based on CA modules, and the elementary traffic vehicles are modeled with Agents. Based on a detailed discussion about the influence of driver' s characters on their driving behaviors, a framework for modeling human-like driving behaviors for vehicle agents is constructed using Fuzzy logic method. Based on this framework, the methods for making and implementing the most important driving behaviors including free-driving, car-following and lane-changing are researched, and basic driving behaviors models for three type vehicle agents are further constructed.Finally, a graphic traffic simulation prototype system based on our method is implemented, and the traffic flow features of simulation are analyzed. The results prove that the new modeling method of traffic simulation is feasible and reliable.