Research Of Traffic Signal Optimization Based On Improved NSGA Ⅱ

Intersection, as an important part of traffic flow scheduling, the traffic signal timing will directly affect the traffic efficiency. The paper found that most scholars usually consider the vehicle performance, ignore the pedestrian benefits. Due to some period, the number of pedestrians will increase sharply, then pedestrians stay the longer, it is the danger, the pedestrians is not negligible.In view of the current traffic congestion phenomenon, this paper established a multi-objective optimization model of traffic signal timing. Given the elite reserved strategy of NSGA Ⅱ make vast redundant high sorting individuals as an elite to the next generation at the same time, to occur premature convergence phenomenon easily, proposes the Ⅰ-NSGA Ⅱ, and apply to the intersection traffic signal optimization problem. Ⅰ-NSGA Ⅱ proposes the method of marking redundant individuals. Later, the elite reserved strategy judges the redundancy individuals through the flag and incorporates it into temporary hierarchy. Last, when the size of new group is insufficent, it will take out proper scale redundant individuals and incorporates them into new group after mutation, in order to improve the population diversity, reduce the phenomenon of premature convergence.According to the problems above, conducted the algorithm experiment. The results show that with the increase of population size, Ⅰ-NSGA Ⅱ won’t produce redundant individuals, reducing premature convergence phenomenon; Ⅰ-NSGA Ⅱ reduces the vehicle and pedestrian delay, and improves the traffic intersection comprehensive transportation efficiency; Pedestrians efficiency will not influence the vehicle performance.