设置排阵式预信号的干线交通信号协调控制优化

Optimization of arterial traffic signal coordinated control with tandem pre-signal

为提升干线道路整体的车流运行效率,建立了一种优化设置排阵式预信号的干线交通信号协调控制系统配时方案的双层模型,并提出对应求解算法;双层模型的上层模型为主信号间相位差优化模型,采用遍历搜索算法优化主信号各交叉口间的相位差;下层模型是以通过车辆数、车均延误为优化目标的多目标优化模型,建立了多目标花朵授粉算法(FPA)对其求解;双层模型中的交通参数通过冲击波建模进行关联,通过上下层模型的迭代求得参数的最优解;以设置排阵式预信号后3个连续交叉口为研究对象,应用提出模型优化高、低2种交通需求下的干线道路交通信号协调配时方案,通过SUMO软件测试所选方案的有效性。研究结果表明:该双层模型能够优化设置排阵式预信号的干线交通信号协调配时方案,与传统干线信号协调控制方案相比,提出方法的配时方案在高、低交通需求下系统通过车辆数可分别增加16%~35%与8%~17%,延误分别降低7%~17%与2%~16%;相较于粒子群优化(PSO)算法与二代非支配排序遗传算法(NSGA-Ⅱ),FPA达到指定精度要求的迭代次数分别减少13和24次。通过仿真结果可知,所提出模型可进一步提升高需求状况下道路的运行效率。

To improve the overall traffic flow efficiency,a bi-level model was established for optimizing the timing plans of arterial traffic signal coordinated control system with tandem pre-signals,and its solving algorithm was proposed.The upper-level model of the bi-level model was an optimization model of the offset between main signals,and the traversal search algorithm was employed to solve it between intersections.The lower-level model was a multi-objective optimization model,which selected the throughput vehicles and the average delay time as the optimization objectives.The flower pollination algorithm(FPA)was established to solve the proposed multi-objective optimization model.The traffic parameters in the bi-level model were connected by using the shockwave model.The optimal solutions of the parameters were obtained through the iterations between the upper-level and lower-level models.Three consecutive intersections after setting up tandem pre-signals were chosen to test.The proposed method was applied to optimize the traffic signal coordination timing plan on arterial roads under both high and low traffic demands.The effectiveness of the selected scheme was tested by software SUMO.Research results indicate that the bi-level model can optimize the arterial traffic signal coordination timing plan with tandem pre-signals.Compared with the traditional arterial signal coordinated control plan,the timing plans obtained from the proposed methods can increase the throughput vehicles through the system by 16%-35%and 8%-17%,respectively.Under high and low traffic demands,and the delays reduce by 7%-17%and 2%-16%,respectively.Compared to the particle swarm optimization(PSO)algorithm and non-dominated sorting genetic algorithm-Ⅱ(NSGA-Ⅱ),the FPA requires 13 and 24 fewer iterations to achieve the specified accuracy requirements,respectively.The simulation results indicate that the proposed model can further improve the operational efficiency of road under high demand conditions.2 tabs,7 figs,30 refs.