| These years the coordination control of regional traffic signal has became a hotspot inintelligent traffic since the problem of traffic jam becomes increasely serious.Most of thesestudies used the delay model in HCM,but there is not phase difference in the model,and someresearches about phase difference didn’t elaborate the relation between phase difference andcycle time,phase time,vehicle speed,distance of two intersections detailedly.So in this paper thedelay model with phase difference will be added for making the intersections being associatedwith each other,and the model is based on different conditions which is concerned with therelation between phase difference and cycle time,phase time,vehicle speed,distance of twointersections.Based on the delay model,genetic algorithm will be used to optimize the cycle time,phasetime and phase difference of each intersection in a region. In genetic algorithm the probabilityof crossover operator and mutation operator should be related with the average fitness of thepopulation and the fitness of the individual for improving the algorithm.In the solving processpartial optimal individuals of every intersection should be combined as the initial populationand the range of cycle time should be limited for increasing the convergence speed,and todivide the individuals into several segments is good for increasing the diversity of theparameters. The target is to minimize the average vehicle delay of the region, and theminimum unit for studying is single traffic flows, for achieve the goal of global control thetotal region must be the objective for coordination control.Finally simulate the plan proposed in the paper under three conditions in a region with fourintersections, and make a comparison to the unimproved plan. The result indicates that the planin this paper can obviously improve the effect of traffic signal coordination control while thetraffic flow level is not closed to the saturated level. |
PDF Full Text Request