摘要
为提高城市干道协调控制的稳定性和效益费用比,将其与多时段控制有机结合,形成符合我国国情的高效控制策略,首先基于干道交叉口关联度模型确定了协调控制的边界范围,进而对需要进行协调的各交叉口历史交通流量进行了混合聚类分析,得到了相应的多时段控制方案。在此基础上,以干道带宽最大和沿线交叉口车均延误最小为目标,建立了干道多时段协调控制的优化模型,并采用多目标粒子群算法对其进行求解,确定了干道协调控制下各交叉口多时段控制方案的最佳切换时刻。仿真结果表明:本研究的控制模型与现状控制方案、交叉口混合聚类模型相比,沿线交叉口车均延误分别降低12.63%和2.45%,干道多时段的总方案带宽分别增加0.98%和23.51%,且方案切换对交通流造成的扰动的可能性降到了最小。
In order to improve the stability and benefit-cost ratio of coordinated control for urban arterial roads,to combine coordinated control with time-of-day control strategy,and to form the effective control strategy suitable to the situation of China,the boundary of coordinated control is determined by the correlation model of arterial road intersections,the mixed clustering analysis on the historical traffic volumes of the intersections to be coordinated is conducted,and the corresponding time-of-day control scheme is obtained.On this basis,with the objective of maximal bandwidth and minimal average vehicle delay of intersections along arterial road,the arterial road time-of-day coordinated control optimization model is established,which is solved by the multiple objective particle swarm optimization,thus the best switching time for time-of-day control scheme of the arterial road intersections under coordinated control. The simulation result indicates that compared to current control scheme and intersection mixed clustering model,the proposed control model can minimize the average vehicle delay of the intersections along the road by 12. 63% and 2. 45% respectively and can increase the bandwidth by 0. 98% and 23. 51% respectively, and can reduce the possible disturbance on traffic caused by switching traffic control scheme as much as possible.
出处
《公路交通科技》
CAS
CSCD
北大核心
2017年第8期114-122,129,共10页
Journal of Highway and Transportation Research and Development
基金
国家自然科学基金项目(61403255)
上海高校青年教师培养资助计划项目(slg12009)
上海理工大学人文社会科学研究基金项目(16HJPY-QN14)
关键词
交通工程
关联度模型
多目标粒子群算法
多时段控制
协调控制
混合聚类
traffic engineering
correlation model
Multi Objective Particle Swarm Optimization(MOPSO)
time-of-day control
coordinated control
mixed clustering