Research Of Single Intersection Traffic Signal Based On Fuzzy Control

Single intersection is a basic element of urban traffic system. It affects the whole urban traffic system. The traditional intersection control methods are relatively mature, but are unsatisfied with the increasingly traffic problems; on the other hand, the practical application of intelligent traffic control method at the intersection is still relatively small, so the research about the single intersection intelligent control is increasingly important.In this thesis, fuzzy control theory is applied to multi-phase single intersection control. The classical fuzzy control method is used to control the single intersection. Simulation based on Matlab program is executed in different traffic flow sconditions, and the validity of the fuzzy algorithm is verified by comparing with the traditional intersection control methods.A two-stage fuzzy control method is used to control the multi-phase single intersection in this thesis. The method also improves the transport phase hopping mode. The improved two-stage fuzzy control method contains three modules:the red light module, the green light module and the decision-making module. The red light module and the green light module process corresponding traffic flow, then the decision-making module handles the results of the above two modules and makes the final decision. By the processing of three modules, the improved two-stage fuzzy control method makes extraction of fuzzy rules easily and reduces the number of fuzzy rules.Then Matlab simulation is made in different traffic flow situations.This thesis presents an improved fuzzy control method by using Genetic Algorithm to optimize the fuzzy rules. It reduces unreasonable rules and improves the efficiency of genetic algorithms. Because of the variability of fuzzy rules, the fuzzy control system has real-time ability. Comparing with the results of two traditional fuzzy control methods, the Matlab simulation in different traffic flow station shows the effectiveness of the improved fuzzy control method.