Traffic Signal Timing Optimization Based On Fuzzy Control

With the rapid development of urban economy, the continuous increasing of population and motor vehicles make urban traffic congestion become more and more serious. However, road intersection is the basic unit of urban traffic road network, and it is also the key place of traffic jam, therefore it is significant to study the optimization of intersection signal timing. To the complexity and fuzzy of the transport system, this thesis uses the fuzzy control theory to optimize the signal timing of intersection.In the single intersection signal control, the single stage fuzzy control is suitable for the low flow traffic state, while the control process of the two-stage fuzzy control is more complex. In order to solve this problem, this paper presents a signal timing method based on combinatorial optimization fuzzy control. The method can select the controller structure based on the traffic states, to realize that the signal phase sequence-changeable and the signal timing is dynamic, also more adapt to the changes of traffic flow conditions. The experiment shows that this method can effectively reduce the average vehicle delay compared with the traditional method during flat peak period and peak period.In the actual traffic signal control, it can not get better control effect only to the timing optimization of single intersection, and also need to coordinate the signal timing of multiple intersections. Therefore, this paper extends the single intersection combinatorial optimization fuzzy control method to trunk multiple intersections, which have the main line direction. In order to reduce the average vehicle delay in the main line direction, this thesis presents a signal timing algorithm that based on the combinatorial coordination fuzzy control for the multiple intersections. This algorithm is used to determine timing parameters of the trunk line by coordination layer and the control layer, and to realize the coordinative optimization of the multiple intersections. Through comparing the algorithm with the general fuzzy control algorithm, the result shows that this algorithm can make the signal timing more reasonable, and it can reduce the average vehicle delay of trunk multiple intersections.To overcome the defect that the primary membership functions can not be changed with the change of traffic conditions in the fuzzy control, this paper uses the genetic algorithm to optimize the membership functions of the green time delay fuzzy controller that in the trunk multiple intersections control unit, also make the membership functions more adapt to the changes of traffic flow conditions. The result shows that the fuzzy control algorithm that through optimizing the membership functions can further reduce the vehicle delay.