单点交叉口鲁棒优化信号配时研究  

Robust Optimal Traffic Signal Timing of Urban Single-point Intersection

作　　者：张萌萌[1,2] 贾磊[1] 邹难[1] 周丽[1] 

机构地区：[1]山东大学控制科学与工程学院,山东济南250061 [2]山东交通学院,山东济南250023

出　　处：《公路交通科技》2011年第1期107-111,共5页Journal of Highway and Transportation Research and Development

基　　金：国家自然科学基金项目(60674062);教育部高等学校博士学科点专项科研基金项目(20060422054);山东省自然科学基金(ZR2009GM032);山东大学自主创新基金(2009TS046)

摘　　要：为了消除单点信号控制不适应交通流波动的缺陷,提高信号控制的稳定性,建立了多目标信号配时优化模型。该模型以平均延误时间最短,通行能力最大,以及鲁棒性最好即流量波动时车辆延误标准差最小为目标,以有效绿灯时间、总时长、各方向最大滞留车辆数为约束条件,对定时信号配时参数进行优化,并利用遗传算法对模型进行求解。求解结果表明,该方法与Webster方法相比,减少了车辆平均延误;与以延误最小、通行能力最大为目标的优化函数相比,缩短了单位时间最大滞留车辆数,减少了车辆最大延误,提高了绿灯利用率,降低了信号控制对流量波动的敏感性。In order to eliminate the deficiency that signal control of single-point intersection cannot meet fluctuations of traffic flow and improve the stability of signal control,a multi-objective signal timing optimization model was established.The object of the model is to minimize the average delay,maximize traffic capacity,and optimize robustness of signal control（minimize standard deviation of vehicle delay）.Constraints of the model are effective green time,cycle time,and the maximum number of stranded vehicles of the directions.The model was used to optimize the signal timing parameters and genetic algorithm was used to solve the model.The result shows that（1） compared with Webster method,the method could reduce the average vehicle delay;（2） compared with the optimization function which taking delay minimized and traffic capacity maximized as objective function,the method could reduce the maximum number of stranded vehicles per unit time,shorten the maximum delay of vehicles,improve utilization of green time,and reduce the sensitivity of signal control to flow fluctuations.

关 键 词：交通工程 信号配时 遗传算法 单点交叉口 鲁棒优化 

分 类 号：U491[交通运输工程—交通运输规划与管理]