降雨天气单点交叉口交通信号控制优化方法
Optimization methods for traffic signal control of isolated intersection under rainy weather

为优化降雨天气下单点交叉口交通控制的效率,引入了降雨修正系数,建立了单点交叉口交通信号配时方法; 应用行动分析软件分析降雨天气交叉口处的高精度视频,以标定饱和流率、损失时间和到达车速等参数的变化特征,提出了饱和流率、损失时间与到达车速的降雨修正系数; 建立了基于模拟退火算法的优化模型,计算了各降雨等级下的修正系数; 构建了基于VISSIM仿真测试环境,评估了提出的模型优化降雨天气下单点交叉口交通信号控制的效果,分别比较了采用优化后参数的定时配时与感应配时对采用原参数的定时配时方案的交通运行效率。分析结果表明:小雨、中雨与大到暴雨天气下的平均车头时距分别比正常天气增加了0.314%、1.256%、2.871%,平均损失时间分别增加了1.042%、2.829%、3.424%; 在流量低于600 pcu？(h？lane)-1时,改进的感应控制方案效果较好,比采用原预设参数方案的车均延误降低了12%～23%; 当流量高于600 pcu？(h？lane)-1时,采用改进的定时控制方案效果较好,车均延误比原方案降低13%～25%,并可在临近饱和与过饱和状态时推迟锁死状态的产生,车均延误最低。
To improve the efficiency of traffic signal control at an isolated intersection under rainy weather, a traffic signal timing method was established by introducing a rainy correction coefficient. The high-resolution video of the isolated intersection under rainy weather was analyzed by using the motion analysis software, the characteristics of the saturation flow rate, lost time, and approaching speed under rainy weather were calibrated, and the rainy correction coefficients of the saturation flow rate, lost time, and approaching speed were established. A optimization model was proposed based on the simulated annealing algorithm, and the correction coefficients under various rainfall intensity levels were calculated. A VISSIM-based simulation environment was built, and the isolated intersection traffic signal control effect of the proposed model was evaluated under rainy weather. The traffic efficiencies of the isolated intersection using pre-timed timing control and actuated timing control with optimized parameters were compared. Analysis result indicates that the average headways under little rain, moderate rain, and heavy or torrential rain increase by 0.314%, 1.256%, and 2.871%, respectively, the average lost times increase by 1.042%, 2.829%, and 3.424%, respectively. When the flow rate is lower than 600 pcu？(h？lane)-1, the improved actuated control exhibits better performance and its average delay is 12%-23% lower than that of the original plan. When the flow rate is higher than 600 pcu？(h？lane)-1, the effect of pre-timed control with the optimization is better, and the average delay is 13%-25% lower than that of the original plan, which can delay the lockout states under the saturated and over-saturated conditions and lead to the lowest delay. 6 tabs, 12 figs, 30 refs