中文摘要：

针对单路口交通流运行的高度复杂性和随机性，本文通过引入演化博弈理论和选择机制，以路口车辆平均延误时长最小为目标，提出了一种新的信号灯控制算法——基于演化博弈的单路口信号灯配时优化控制（EGSTOA）．为了评估方法的有效性，在VISSIM仿真软件中进行实验，并将该方法的控制效果与优化定时控制和基于遗传算法的控制效果进行比较．对比结果表明，使用EGSTOA方法有效降低了路口车辆平均延迟时长，改善了路口的控制效果．

英文摘要：

Aiming at the eomplex and random traffic flow at single intersection, this paper adopts the idea and the selection mechanism of evolutionary game lheory to examine an new signal ligh! eontrol mechanismAn evolulionary game based optimization method on signal timing for single intersection （EGSTOA） , which is made to minimize the mean delay of traffic flow. To evaluate the proposed method, the performance of this method is evaluated by the traffie simulation software VISSIM, thereafter, this study compares the control effects of this new approach with the fixed-time control method and the signal control based on the genetic algorithm. The simulation result illustrates that the EGSTOA method effectively reduces the mean delay of traffic flow and improves the effectiveness of signal control.