Research On Urban Mixed Intersections Traffic Flow By Cellular Automaton

Traffic congestion is a key problem for the modern society developing. The arterial road is the main part of urban traffic. And it is also the most significant influences of the city traffic. Therefore, in order to effectively solve the traffic problems, improve vehicle's efficiency, and make the road flow in good condition, it is necessary to make the scientific analysis for traffic conditions, and find effective measures.The cellular automata (CA) were given a deep research. First of all, we establish a simulation model of traffic road based on the theory of cellular automata. Simulations and theory analysis prove that CA has correct dynamics characteristic of traffic flow. Secondly, according to actual road condition, the model of arterial road intersections was established. The effects of different intersections and structures on traffic flow have been simulated and analyzed. Thirdly, the mixed intersection model was established, an improving algorithm of traffic lights control was proposed. This can make the roads traffic flow better, and vehicles move more smoothly through each intersection of the arterial road. Then, according to the fuzzy neural network theory, we put forward a control algorithm of traffic lights which based on the cellular automata combined fuzzy neural network. The simulation results show that the traffic flow which in control of fuzzy neural network is much better than synchronous traffic light. Finally, the influence of two adjacent arterial roads was analyzed, and a new algorithm of turn right probability which based on fuzzy neural networks was presented. This algorithm can make the model more realistic. Comparing the results before, we find the adjacent arterial road can shunt vehicles, make congestion fast dispersed.The model of arterial road intersections was established and the relationship among density and speed, and traffic flow of the arterial road in different cases were analyzed. For the mixed intersection, several innovative ideas have been presented. (1) An improved algorithm of adaptively changing the green time was proposed. (2) A harmonious control model to adjust the traffic light distribution is put forward based on the improved adaptive algorithm combined with fuzzy neural network. And (3) a new traffic model considering the condition of the adjacent arterial road was given. And a new algorithm of turn right probability based on fuzzy neural network. These modifications make the proposed model fit the real. The simulation results and theory analysis show that these methods are propitious to the traffic.