中文摘要：

针对棋盘状城市路网交通信号群，提出一种基于饱和度阵列平滑的信号配时算法，并且建立一个路网配时优化系统．采用两种二维模板分别对整体路网信号周期和绿信比进行调控，其中拥堵程度评估模板从小区域进行拥堵判定并对中心交叉口调整周期时长；拥堵方向四邻模板从相邻交叉口空位分析中判断绿信调整方向．从面到点进行宏观到微观的调控，使路网信号群协同工作．动态阈值使路网处于对车流速度能产生线性影响的正常饱和区域，空域滤波使整个路网饱和度快速平滑和降低，进而提高了整体车流的行驶效率，间接降低了整体车辆延误．仿真结果表明：饱和度平滑优化方法效果显著优于传统静态方法，适合于突发性或整体严重不均衡的交通状况，也适合于高饱和度交通状况．

英文摘要：

This paper proposed an algorithm based on saturation smoothing and built an optimization system of the network timing that two kinds of two-dimension template respectively control the network signal cycle and green ratio on the checkerboard road network. The template of degree evaluation determines the degree of congestion from a region and adjusts the cycle length of center intersection signals. The template of direct evaluation decides the direct of green signal. Macro to micro regulation makes the network signal group cooperative work. The road network works in a linear area of vehicle speed to saturation by a dynamic threshold. The spatial filtering smoothed and reduced the network saturation, so to improve the whole traffic flow efficiency and reduce vehicle delay indirectly. The simulation results show that the effect of saturation smoothing optimization method significantly outperforms the traditional static method. The method is suitable both for the bursty or whole seriously unbalanced traffic conditions, and for the high saturation traffic conditions.