Abstract: The paper presents a fuzzy control system structure for an urban junction adopting fixed phase orders and multiphase. The traffic flows in the next cycle may be estimated, and the maximum time of green light is adjusted dynamically. Each phase time is determined according to traffic urgency degree of current phase and next phase. Using genetic algorithm to modify fuzzy membership functions, the system can change its membership functions adaptively to an optimal setting in different traffic situations. A simulation experiment at intersection is performed and the performance is measured with average vehicle delays through the intersection. Simulations result shows that the fuzzy control system can effectively decrease the average delaying time, improve traffic average velocity in the intersection and is superior to the conventional method.

Key words: traffic flow, multiphase, fuzzy control, genetic algorithms, optimization

摘要： 针对城市固定相序的单点交叉口多相位交通信号进行控制，设计了基于车流量预测的动态调整相位最大绿灯时间的模糊控制系统。综合评估当前相位、后续相位的交通需求度，以此决定绿灯时间分配。系统采用遗传算法对模糊隶属度函数进行优化调整，使隶属函数的选取更为合理，随交通状况的改变自适应地调整。仿真结果表明，该方法能有效降低通行车辆在交叉口的平均等待时间，提高平均车流速度，控制效果明显优于传统控制方法。

关键词: 交通流, 多相位, 模糊控制, 遗传算法, 优化