单点交叉口信号灯优化配时研究

提出了单点交叉口信号灯优化控制新算法，通过对单点交叉口每个方向每条车道未来时刻车辆到达数的有效预测，采用模糊控制推理信号灯下一个周期的时长．在此基础上，建立以车辆平均延误最小和交通流量最大为目标的多目标优化模型，采用排序-遗传算法优化得到每个相位的绿信比，根据此绿信比将周期时长分配到各个相位，从而得到了最优的信号灯配时．仿真试验表明该算法是有效的．

Optimal time assignment of traffic signal cycle in isolated intersection

An original optimal algorithm was developed for controlling the signal in isolated intersection. The time length of the next cycle is obtained by fuzzy reasoning based on the effective prediction of the arriving vehicles in every direction and every line in the next cycle, The multi-objective, optimal model is built to get the least time delay and the most traffic flow capacity, The objective function was optimized by sorting-GA, The split of each phase is obtained, The time length is divided into each phase according to the split. So the optimal signal time assignment is completed. By the computer simulation, the result shows that the algorithm is effective.