基于改进遗传算法的交叉口模糊控制研究

为了改善城市道路交叉口交通信号控制，降低交叉口车辆延误，提出了一种基于改进遗传算法优化的模糊控制方法。在优化模糊控制器的过程中，为避免出现“早熟”现象，采用改进的自适应遗传算法，在进化过程中动态调整种群中适应度值最大个体的交叉概率和变异概率，使种群进化不会处于一种近似停滞不前的状态。为了检验该控制方法的性能，以交叉口车辆平均延误作为性能评价指标，在相同交通条件下进行了仿真实验。结果表明，相对于普通模糊控制器，经过改进遗传算法优化的模糊控制器能有效减小交叉口车辆的平均延误，提高交叉口的通行能力。

Study of fuzzy control for intersections based on improved genetic algorithm

In order to decreas vehicle delay in intersection, this paper presented an improved GA-based fuzzy control method for the traffic signal in urban intersections. To avoid prematurity problem, applied improved adaptive GA to optimizing the fuzzy controller. The crossing and mutation probability of individual that enjoyed the maximum fitness value were dynamically adjusted during the evolution of the population, which ensured that the evolution of the population will not suspend. To validate the performance of this control method, carried out simulation with vehicle average delay in the intersections being the performance index. The result indicates that the optimized fuzzy controller outperforms the traditional one in decreasing the vehicle average delay and increasing traffic capacity of the intersections.